Scientific Observation : Image and Reality

Philosophy Forums Science, Technology, and Math
1

Here’s an article I’ve been attempting to better understand. The topic is about a distinction between lensed and unlensed observation. I sort of doubt that many people will have the time or interest to study and discuss it. But if someone should, an interesting conversation might emerge. Please refrain from one line responses that do not pertain to the subject. I’ll be going back over this article in detail in an attempt to understand all the information within it. I think it has a great deal to do with science and epistemology. If that’s interesting to you, and you’ve got the time and patience, let me know and I’ll start brushing up on it and we may just end up with an newer understanding of what’s probably an old problem in science and observation. Good luck guys. I look forward to others rising to the challenge of understanding this piece well enough for all of us to understand a new perspective on this matter. Thanks for your interest. Let me know what you think.

-Smears

2

ABSTRACT

On the standard Source-Transmission-Receiver (STR) philosophical analyses of observation (e.g., Shapere and Paller), a receiver observes a source just if the source transmits a signal directly to the receiver, and the receiver receives the signal. There are two problems with this account: first, observation of an effect is indistinguishable from observation of a source; and, second, there is no good way to distinguish between an observation and other causal chains.

This dissertation develops an alternative philosophical account of observation that resolves both difficulties. It distinguishes two kinds of observation: lensed observation (e.g., via eyes), which is the result of an inverse Fourier transform applied to radiation which interacts with an object, and is transmitted to a receiver; and unlensed observation which does not use a transform of this kind to detect effects.

The account offers a resolution to the first problem with STR analyses by showing that in lensed observation what is observed is the object with which electromagnetic radiation interacted, and to which an inverse Fourier transform is applied, not just effects of the object. In unlensed observation, what is observed are just effects.

The account also offers a resolution to the second problem with STR accounts. The solution requires that observation be observation of something, and so requires individuation (on the basis of the inverse Fourier transform or a theory of the instrument) of an object or property. This means that not all interactions in nature are observations: Only those in which interaction also results in individuation of an object or objects are observations.

The solution does not presuppose that all observations are either lensed or unlensed.

Because these two kinds of observation have significantly different epistemic properties, sorting them out is essential to any satisfactory philosophical analysis of observation. However, most accounts of standard problems on observation in the philosophical literature, including those of van Fraassen, Fodor, and Hacking, do not distinguish these two kinds of observation and encounter difficulties as a result. Many of the most controversial claims of these philosophers are confusions rooted in failure to sort out the different philosophical properties of lensed vs. unlensed observations. I argue that the account developed here offers a better philosophical understanding of observation while resolving these problems.

3

I’ll post the rest if anyone finds this interesting. Also, the author is Sara Vollmer, just to make sure she gets her credit. Thanks guys. Someone blow my mind and take interest in this. I’m pretty pessimistic about that actually happeneing. It seems like lately the philosophy forum has become a self help group for upset teenagers. If you’re one of them, why not go the fuck away and find some other place to make spurrious arguments against common sense? If you’re not, then why not take interest in some real philosophy rather than the dry and boring history of it which most of the time leads nowhere and adds no new information for the rest of the people here who have an interest in learning. Thanks to all of those who are interested.
-Smears

4

Hi Smears,

The following diagram is how I see the classical model of scientific observation. This is exclusive of the dasian and my consciousness located at the far left and far right of this image.

Could you please explain: What the Source-Transmission-Receiver is in this picture?

I have no idea what the inverse Fourier transform has to do with any of this. (I’m pretty good at this stuff)

Thanks Ed

P.S. Lots of things can and do go wrong here!

5

My assumption is that the source-transmission-reciever there would be the ball, (source), the radiation, (transmission), and the optic nerve,(reciever). Maybe I’m wrong here. I’m really not that good at this stuff, but I think the distinction that she’s making as far as the fourier transform goes has to do with the distinction between lensed and unlensed observations. If something is transmitted via fourier transform, then what you’ll get is an exact rendering of the partcular spectrum of reflected light or what have you from the object onto the lens. I’ve got a long way to go before I can fully discuss this. I hope this helps. If you think I’m wrong about something I’m totally willing to read and learn. Thanks for responding.

6

I’m pretty skeptical on this. Seems to be some kind of pseudo-science rant that makes very little sense and is rather pretensions. But maybe I just don’t get it. Why should we distiguish between observation with the eyes compared with other observations? I really don’t see why lenses make that much difference. It sounds very interesting but I honestly think its bullshit.

7

Hi Smears,

Source, Transmission, and Receiver Comments

I, too, initially, thought that the Ball, Radiation, and Optical Nerve served the purpose of the Source, Transmission, and Receiver respectively.

However, if we were to interpret this broadly, we could consider any three contiguous transition places as the Source, Transmission, and Receiver. And, more broadly yet, we could include any four or more contiguous transition places where the transition places between the source and the receiver were considered as a composite transition. (This would be analogous to composite functions.)

Finally, I am curious about the roll of the receiver. For example, if the receiver simply transmits data, and no one sees this data, is it not analogous to a tree falling in the forest and no one hearing it? Even if it preserves the data, without some sort of pattern recognition would not the data simply be the equivalent to statistical noise – basically meaningless? I think that the only receiver of consequence is the observer/brain.

From this perspective the lens plays no particularly meaningful role; it is simply part of the transmission.

Tweaking the Scientific Observation Model

In the diagram, I show an individual observing a ball. Science, and perhaps common sense, tells us that cultural and personal bias will affect what we see.

The inversion of the old expression “I’ll believe it when I see it” to “I’ll see it when I believe it” is actually true in some cases. Alan Chalmers book “What Is This Thing Called Science” gives some good examples.

This means that better science avoids anecdotal evidence, and requires public viewing.
In the case of Astronomy photographic plates are probably thought of as the receiver and this captures the data. Groups of people can then view the data at a later time and a collective understanding emerges. I would note that the plates themselves do not convey meaning. It is only after they are viewed that meaning is actually conveyed.

Finally science does appreciate an enhancement of our senses. Bigger and better lens appear to give us better detail.

With regard to the inverse Fourier transform

The Newtonian world is effectively based on the equation F = ma. Since a is a vector acceleration, we are dealing with differential equations. In order to model certain types of behaviour, particularly strings and heat transfer, these equations become very complex, and require the use of partial differential equations. (Many of the component forces are actually determined by trail and error).

Probably the most useful tool in dealing with these equations is the Fourier transform. (Though there are other transforms e.g. the Laplace transform.) What happens, in some circumstances, is that the Fourier transform changes complex partial differential equations, in the standard domain, into simpler equations in this new “Fourier Domain”. These simpler equations can, in these special circumstances, be solved in this “Fourier Domain”.

However, for us to make use the solution in the “Fourier Domain”, we need to change the solution back to its standard domain. This transform to go back is called the inverse Fourier transform. (The solution to this transform does not always exist and people have tweaked the integral used in the definition of the inverse Fourier transform to change the class of functions, in the “Fourier Domain” which can have inverse Fourier transforms. The best known of these people is Henri Lebesgue, but even I have played with this integral definition).

How might this apply?

I am not sure. It is possible that any lens can interact with the radiation, both through heat distortion and possibly, though minimally under ordinary circumstances, exchange of energy with the radiation which, in theory, could cause a vibration.

However, the solutions to the Fourier Transform problem are going to be heavily dependent on the modelling in the first place and the math is basically just an after thought. If you think of this in terms of Chaos Theory, the solutions are generally no where close to linear and the initial conditions are highly dependent on the modelling.

Additionally the same problems would occur with an optic nerve as would occur with a lens, yet no discussion of this matter has taken place.

Do your authors give you any more information about the specifics of the circumstances and their analysis of the variations caused by lens?

Anyway such are the thoughts of a senile old man, who has not even read or listened to the body of work you are presenting.

8

I’d like to read the rest before I weigh in. Keep up the good work.

9

If I read this right, I see an erroneous assumption, and a ‘hypothesis’ built upon it. A ‘house of cards’.
The erroneous assumption? That of the Perspective of ‘linearity’ (as ‘Reality’) and the simultaneously arising feature of Perspective, ‘cause and effect’ (scientifically obsolete).
These notions only have any ‘validity’ in the ‘reality’ of a certain few distinct Perspectives (us). This ‘appearance’ does not span all Perspectives and are therefore not ‘universal’ (as ‘perspectival appearance’ suggests).
All Perspectives are ‘correct’ within that Perspective (for you). All Perspectives are incomplete (to some degree). ‘Linearity and c&e’ are far from ‘universal Truth’, other than a ‘truth’ in iour[/i] universe.
The act of Perceiving/conceiving mutually arises with (y)our existence/the existence of ‘(y)our’ universes. What you Perceive is a unique Perspective that is your world. Perception cannot be seperated from existence. What you ‘see’ is correct. What you ‘think’ about it is (generally erroneous from a 'scientific view) is consequential to ‘your’ Perspective.
I think that this relates to the OP (‘odd’ as it is)…
Peace

10

Hey Smears, thanks for posting this. In my post here I criticize the distinction Vollmer’s abstract attempts to define, and below I suggest what the distinction really should be. Perhaps in the main text of her thesis she is more clear, but as it is I think her language and conceptual boundaries need improvement.

What are “observation of an effect” and “observation of a source”, such that Vollmer can claim they are indistinguishable? Usually when we talk about “observing a source”, what we mean is “observing an effect followed by inferring its source”. We see blue light in a certain pattern, then infer that there’s a ball in front of us – these two together make the act of observing the ball. So I would say “observation of an effect” and “observation of a source” are quite distinguishable. One is part of the other.

Second, Vollmer says we can’t distinguish between observation and other causal chains. I find this very strange. Suppose I set up a very simple experiment to measure how much light is being transmitted to a certain spot on the wall. I shine a light at the spot, the light hits a detector in front of the spot, the detector shows me a value, say 10 Watts. To me, the observation is clearly my act of reading and recording the number. How is it so hard to distinguish between the observation and the other “causal chains” happening in this experiment? (But see my comments below – sometimes scientists consider the reception of the light by the detector to be the “observation”. In either case, the observation is an observation because the experimenter declared it to be one.)

Given what I’ve just said, to me this distinction makes so sense. To me, all observations are through eyes, and it’s not of any particular importance what Fourier transforms may or may not happen. (But see below, where I have a guess on what Vollmer is really after here.)

I think of the STR model mathematically. If the observation is y, the source is x, and f is the process (the “stuff that happens”) between the observation and the source, we have

y = f(x).

In the STR model, we have
S = x
T = f
R = y.

To observe an effect means to mentally record y. To observe a source means to record y, then infer x by some sort of mental analysis of f. This analysis may occur unconsciously and automatically, as when we observe red balls “directly”; or it may require deliberate thinking and mathematical sophistication, as when we observe elementary particles.

However, sometimes the word “observation” is used in scientific circles slightly differently than I am using it here. (Perhaps Vollmer recognizes this too and is attempting to tease apart the different meanings.) In this slightly different usage, observation can be done by a non-subject like a spectrometer. For example, I once ran an experiment where I had a computerized spectrometer “observe” the reflectance of a sample vs. wavelength of incoming light. (Light came from a source, hit a sample, then was partially reflected into a receiver.) In this case an “observation” is an “endpoint of modeling”, a point in physical space and time when I claim that my experiment has created the information I want in the form that I want it (whether I have observed this information or not!)

For example, let’s think about this reflectance experiment. When I set up my y=f(x) equation, I let the light source be x; I let the optical dynamics be f; and y is the data the computer records, stores, and displays for me to read. But what about the wiring connecting the recorder to the computer hard drive and display? Those are physical effects just like the optical dynamics are, and in some sense they’re part of my experiment. But in another sense they’re not, because I’m not going to model them explicitly. I’m going to assume the computer recorded and displayed the right numbers unless I have reason to think otherwise. So for shorthand I say that my spectrometer generated observations of reflectance (perhaps Vollmer would call these “unlensed observations”). In fact this shorthand is a sort of reader’s aid, a rhetorical focusing tool. When I write in my paper that my spectrometer observed reflectance, I am defining an endpoint of modeling for the reader. I push the computer’s workings under the rug and ask the reader to focus on the optical part of what happened.

Thus, I think the real distinction to be made is between observation as a subjective act (subjective-observation) and observation as a rhetorical endpoint of modeling (endpoint-observation). The first is Vollmer’s “lensed” observation and the second is “unlensed”. When we subjectively-observe, we mentally grasp our data y; at this point we may mentally begin the process of inferring x, and hence “observing a source” in my usage at the beginning of this post. When we endpoint-observe, we are merely saying that at this point in the experiment, we have the information we wanted and can stop modeling. No mental reception or interpretation of the data itself (subjective/lensed observation) has necessarily occurred yet, so only the uninterpreted data y can be said to exist – only the effect y has been “observed”, not the source x.

11

A lot of this is over my head, but I’m confused about the receiver part of this. Using a little of Ed3’s model, and comments by aporia, it seems to me that the lensing is merely part of the aparatus to apprehend the raw data. The actual receiver is the brain which interprets the raw data. I too have difficulty accepting the notion that there is a distinction between lensed and unlensed, or that it is even relevent. There are other senses that process information as well. We hear, we touch, we taste, and all of these can fit into Ed3’s model. The critical element is the receiver, which in this case makes sense of all data received from neuronal activity. The multiple sources (radiation) that can trigger neurons firing can indeed relay raw data, but it is the brain that is the actual observer.

Aporia, I agree that sources are inferred. Quantum particle physics has never seen any of the sub-atomic particles. Their existence (as source) can only be inferred from the photographic plates of the cloud chambers, and again this requires interpretation(brain/receiver) . Something is missing here… probably my poor understanding.