I understand the concept of how all observations are theory laden and how it works as a critique of the positivist program. From this link:

Theory-ladenness of observation holds that everything one observes is interpreted through a prior understanding of other theories and concepts. Whenever we describe observations, we are constantly utilizing terms and measurements that our society has adopted. Therefore, it would be impossible for someone else to understand these observations if they are unfamiliar with, or disagree with, the theories that these terms come from.

But I am struggling with how it can extend to any observation.

How can very basic statements like:

  • All objects fall to the ground.
  • Ice melts in the sun.
  • All people die eventually.

Be theory laden? Are these types of observations exempt?

  • what is e.g. melting? is it related to a prior understanding of another theory or concept?
    – innisfree
    Commented Jul 4, 2015 at 21:02

7 Answers 7


You are probably assuming that the statement is true without having to posit ludicrous assumptions. But let's just take the statement at face value, and not make any "sensible" assumptions: "All objects fall to the ground."

What is an object? What is it to fall? What is the ground? You need to have some sort of framework for those things to even interpret the sentence. If you have somehow arranged things so as to have no concept of "fall"--not even any good way to translate it--then, indeed, the statement can't even be interpreted.

This does not mean one should take this as a serious criticism any more than one should take the Evil Demon as a serious argument that all empirical knowledge is bankrupt. It would be very difficult, for instance, to survive as a human without any concept of something akin to "fall" (or "object", or "ground").

As far as I can tell, the anti-Evil Demon manoeuvre (which basically involves the pragmatic decision to ignore the argument) hasn't been sufficiently vigorously explored with regards to logical positivism. Note that empirical scientists, to great success, have basically taken this pragmatic approach; Kuhn* notwithstanding, issues of theory-laden perspective even in case of revolution basically don't come up in the harder sciences because of how strongly constrained the entire theoretical framework is by all the data we've collected.

For less data-constrained sciences (e.g. ecology), it's less clear that theory-ladenness is not a grave practical as well as theoretical concern.

*I credit Kuhn rather than Popper because although Popper was concerned with theory-ladenness of observations, it was Kuhn's work in The Structure of Scientific Revolutions that really made it an essential aspect of the scientific process rather than a minor annoyance to circumvent.

  • I am surprised that you do not credit Quine on linguistic/factual distinction in logical positivism. And he explored a very pragmatic question: given a scientific framework how do we distinguish one from the other? This can not be ignored because it is not a criticism of what is being done, we need an actual solution and none has been forthcoming despite lifelong efforts by Carnap and others. It should be pragmatic, but it has to work in general, not as a series of consensus declarations on what is factual. Because then we get something like universal grammar, and no agreed upon facts.
    – Conifold
    Commented Jul 5, 2015 at 22:48
  • @Conifold - That's highly relevant, I agree (naturalized epistemology especially). But the real puzzle to me is why most philosophers agree that there is knowledge (despite the absence of an actual solution to that either) but disagree that there is effectively theory-unladen observation, even though the underlying logical concerns are similar, and our everyday experience argues strongly for both knowledge and effectively theory-unladen observation. That we can't say quite how we manage a particular analytic/synthetic distinction should not necessarily void all evidence that we do make it.
    – Rex Kerr
    Commented Jul 6, 2015 at 1:46
  • That's because Cartesian demon is contrived but Quine is speaking from experience:"we are now impressed with how stubbornly the distinction between analytic and synthetic has resisted any straightforward drawing. I am impressed also... with how baffling the problem has always been of arriving at any explicit theory of the empirical confirmation of a synthetic statement... Taken collectively, science has its double dependence upon language and experience; but this duality is not significantly traceable into the statements of science taken one by one".
    – Conifold
    Commented Jul 6, 2015 at 21:28
  • Friedman has interesting ideas in Dynamics of Reason on stratifying science into constitutive, coordinating and empirical parts, something of a merge of Carnap's frameworks and Kuhn's paradigms in opposition to Quine's holism. But his 'analytic/synthetic distinction' is different, empirical parts are not naively factual, they explicitly depend on the other two to be meaningful and testable one by one. Constitutive and coordinating parts are not directly testable at all, they stand or fall wholesale. But when they are fixed in place tradition and habit create the illusion of empirical facts.
    – Conifold
    Commented Jul 6, 2015 at 21:30
  • @Conifold - Sounds interesting; I'm not really familiar with Friedman's work. Maybe he's already done what I found missing? But I disagree that the Cartesian demon is contrived in a way that the analytic/synthetic distinction is not. Learning things from introspection is often much like wrestling with that demon, so greatly does our subjective sense of experience try to align well with objective facts while papering over implementation details. That you cannot get everything you want from some framework does not mean that you can get nothing.
    – Rex Kerr
    Commented Jul 6, 2015 at 21:38

Considering your specific question, theory-ladenness applies specifically to scientific observations. If we are to identify theory-ladenness in an observation such as "Ice melts in the sun", then we need first to express this observation using scientific terms - e.g., H2O molecules undergo a phase transition from solid to liquid at a temperature above zero degrees C. (Note that I have modified your observations since if it is minus 50C, then ice won't melt whether the sun is out or not.)

The theory-ladenness of this observation is now evident, since interpretation requires an understanding of molecular theory, temperature, etc..

More generally, it is not clear that the quoted description of theory-ladenness by Adam White is well-expressed. For example, it does not account for the case where an observation is made that falsifies (all) existing theory. In 1777, Antoine Lavoisier's experiments with oxygen led to observations that falsified the prevailing phlogiston theory. The absence of alternative theory meant that these observations could not be interpreted, leading to the need for new theories to be formulated.

It may be that theory-ladenness is intended to apply only to post-19th century science. Prior to the 20th century, science was observation driven. We did our experiments, made our observations, and then formulated a theory to fit the observations. This changed in the 20th century when science became theory driven. We now formulate our theories and then make observations to confirm (or falsify) our theories.

Whatever the case, the problem of an uninterpretable, falsifying observation remains if we are to use the quoted description.

  • Very little of 20th century science is really theory-driven in a way that 19th century science was not. There's a bit of this in high-energy physics and a few other areas, but mostly we still make observations and then try to figure out what they mean. But since we know a lot already, there's no point observing things where we already know the answer, so we try to target it to areas where we are unsure whether we know or not.
    – Rex Kerr
    Commented Jul 5, 2015 at 3:20
  • 1
    @RexKerr You're right, course. The labs are still very active. I guess I was thinking of contemporary physics as primarily mathematical physics. Our mathematical physics such as string theory and relativity are popular and the theoretical physics hogs all of the grant money - e.g., high energy physics.
    – nwr
    Commented Jul 5, 2015 at 3:31

Your three examples are not observations in the primary sense, they are conclusions. "All objects fall to the ground" is a conclusion about all past and future states of affairs of a particular type, and most of those events have not been observed. An observation would be e.g. "This brick fell to the ground when I released it 5 minutes ago". Some observations are clearly theory-laden, such as any measurement of temperature, pressure and acceleration. Other observations are not actually theory laden, but are standard-laden (such as measurement of distance with a ruler).

There is a lot of confusion out there over the distinction between "observation" and "conclusion". An "observation" is a simple perceptual act. You can legitimately observe an object falling to the ground, but when you observe that at time T1 an object is on the table and at time T2 it is on the ground, you cannot claim to have "observed" the object falling.

The statement that "everything one observes is interpreted through a prior understanding of other theories and concepts" is true by dint of the meaning of "interpret" and the nature of "concepts" (we need not even invoke "theories"). To say that a perceptual fact is "interpreted" (e.g. when you "interpret" what happens in your brain's data-processing centers as you see me let go of a brick) is to relate the sensations that strike your mind to other, similar sensations. Perceiving that there is a brick depends on there existing (and you knowing) a concept "brick".

Suppose there were a being that had no concepts at all (such as an earthworm). What would it mean for an earthworm to "interpret" an observation?


I can find instances where you 3 statements are false. Therefore, to make sure they are true, certain conditions, properties, definitions must be assumed/presumed, causing them to be "theory-laden."

Hydrogen gas (an object) falls "away" from the ground.
Ice that is constantly cooled, does not melt in the sun(light).
People (Plato, Newton, etc.) that have achieved "literary immortality" will never die.


You asked how the following examples of observation could be interpreted as theory laden:

(1) All objects fall to the ground.

(2) Ice melts in the sun.

(3) All people die eventually.

Observation (1) is false. There are lots of objects that will never fall to the ground because they achieved escape velocity. But let's consider an object that does in fact fall to the ground. You have to identify the object in question as being the same object over time. There are some physical systems that don't have identity over time. For example, if two electrons interact there is in general no fact of the matter about which electron after the interaction corresponds to which electron after the interaction. You have to understand the idea of a non-zero distance between two objects. You have to measure that distance in some way, and there are some ways you could do that which would be wrong, e.g. - if the ground and the object overlap in your field of view they may not be in contact.

(2) could also be false. You might have the ice in a transparent freezer. But in the case where it does melt your observations still involve theory. It assumes that there is such a thing as ice and that different ice samples are made of the same stuff as other ice samples. Then you have to understand the idea that ice and the water it melts into are the same substance. And again, you need to have some way of determining that the ice has changed phase and that requires some explanation of you make that determination.

(3) may also be false. The human body is a biological machine. There is no reason to think it is impossible to maintain the body indefinitely with the right technology. But in any case. In order to identify the person who was alive at one time with the person who died at some later time, you must have the idea of identity of a person over time. You must also have a way of checking that the dead person and formerly alive person are the same person, which can be difficult.

See "Logic of Scientific Discovery" by Popper especially Chapter 5, and "The Beginning of Infinity", by David Deutsch Chapters 1,2 and 10.


There are no examples exempt from this. What you talk of here is called perceptual theory‐ladenness. You probably came in the know-how of the term through philosophers like N.R Hanson or Pierre Humen. Mind you, almost all of the philosophers that backed this up were idealists and propounded, in their works in epistemology, that the mind is severely limited by past experiences and acquired knowledge.

Take the classic example of trying to think of a colour that you haven't seen before. You can't, right? Here's a reddit thread that should put it in perspective for you.

Kant summed it up pretty nicely when he said that All thought must, directly or indirectly, by way of certain characters, relate ultimately to intuitions, and therefore, with us, to sensibility, because in no other way can an object be given to us. What he meant by this was that humans acted, consciously or subconsciously, on impulses and that to a good extent, it is the guiding power behind our conduct.

Since observation is restricted to past knowledge, we must understand the scope of these terms.

Take for example Ice melts in the sun. Now, each of the words here can be interpreted and understood in a plethora of ways. Here are a few that can be subject to debate:

  1. What all do you consider ice and which of them do you take into account? See this.
  2. When do you consider ice to have been melted? For instance, one could argue that even though it has changed its state of matter, it still has the same constituents of the cube of ice and because it is capable of going back to it, it never did undergo a change.

That a stone falls to the ground is an empirical fact; that it falls to the ground because of some vita is theory-laden; as is it falls because of gravity.

Bare facts, in a sense are exempt; except that we do not normally work without a theory or weltenschaung (worldview); and it takes a serious effort to not do so.

ie Stones fall because of gravity, ice melts because of the sun.

One could say that a stone falls is a 'bare fact'; but a different attitude is to say one has an intuitive (or phenomenological) understanding of gravity; if I throw a stone up, it comes down; it follows a particular shape; coffee poured out into a cup doesn't fall in a 'U' shape etc.

Gravity, as a concept, is generally acknowledged to have been discovered by Newton; but if one looks in Aristotles Physics one notices that he has a theory of gravity too, which following his metaphysics is a contrary; this gives to me at least the sense that the phenomenalogical understanding of 'gravity' is in a sense the originary point of all theories of gravity.

There is another possibility - and that is in Kants Philosophy where the categories structure our sensibility; two of them, space and time are the conditions for our experience; others such as quantity structure how we 'see' things: I see several cups before...

  • For most of human history people had no problem accepting the fact of objects falling without a theory of gravity. Commented Jul 7, 2015 at 3:03
  • 1
    @king: I've edited my answer. Commented Jul 7, 2015 at 23:29

You must log in to answer this question.

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