Ljubomir,

Perhaps you will agree that in an exact science such as biology, subjectivity comes into play for a biologist in making choices of species-of-interest and in choosing locations for field work that are deemed promising. Again, for example, in an exact science such as astonomy, subjectivity comes into play in terms of choices of regions-of-interest (e.g., galaxies or solar systems) to study.

Good question!

Subjectivity is present in every field of knowledge. Why? In each case, subjectivity is part of the picture of each field of knowledge in terms of chosen variables (from a mathematical perspective) and in terms of topics chosen (from a non-mathematical perspective).

Here are some examples:

History: events chosen for exposition.

Palaeontology: era chosen for study of fossils.

Philosophy: questions raised and topics chosen for exploration.

Mathematics: chosen threads and theorems found.

For a discussion of physics relative to subjectivity and fuzziness, see Section 3 of the attached pdf file.

Good question. Subjectivity comes into play anytime you have a judgment based on something other than facts; so, when you are determining whether or not something is right or wrong based beliefs could be viewed as a subjective process.

All fields involve some sort of judgement call when doing research. However, in science your judgment is based on data.

Non-science subjects are associated with subjectivity.

You might say otherwise that subjectivity is not affiliated with the exact sciences!

Ljubomir,

Perhaps you will agree that in an exact science such as biology, subjectivity comes into play for a biologist in making choices of species-of-interest and in choosing locations for field work that are deemed promising. Again, for example, in an exact science such as astonomy, subjectivity comes into play in terms of choices of regions-of-interest (e.g., galaxies or solar systems) to study.

I quite agree with James, subjectivity is present in every field of knowledge; not because it is an inherent component of the field, but as a result of human interaction with the field. Even the definition and demarcation of the subject field is subjective, and in many cases spoils science as the boundaries are subjective and synthetic, and oftentimes abused to support some agenda. Most of what we think are precise and deterministic, is in fact fuzzy at some level, when we hit barriers of knowledge, understanding, instrumental accuracy etc. So, we remain confronted with a delicate mixture of subjectivity and fuzziness, sometimes at a higher level than we would like!

Ludwig,

Excellent post! You have brought out some of the subtleties, identifying cases where subjectivity intermingles with fuzziness in every field of knowledge.

IMHO the meaning of the terms 'subjective' and 'subjectivity' is itself fuzzy.

Linguistically, both terms derinve from 'subject' meaning 'acting person' or 'person performing a activity'. Thus, in the context of knowledge systems, intrinsically it refers to human beings as the originators of some kind of knowledge . For that reason, the knowledge in question may partly differ from one person to another.

Some people are afraid of such a 'relativistic knowledge', but frankly we can't live without it. We can't have exclusively objective knowledge, which is the declared opposite of subjective knowledge: knowledge what is exactly the same for all stakeholders, e.g. members of a collaborative community like society, science, etc. Would be nice, but I guess, you would only have tautologies, which is very meager and abstract. There is, however, nothing inherently bad with subjectivity and nothing inherently good with objectivity. They are just two poles of a continuum.

Having said this, it is a good question to ask, how you might or should 'model' subjectivity. Again, there us not a single answer. Linguistics deals with the question, and offers some models. Philosophy deals with the question, and offers its own models. And of course logic deals with the concept, especially through mamy-valued logics, of which fuzzy logic à la Hájek is the most recent expression - and very successful!

Recommended literature: Cintula et al, 2011, Understanding Vagueness, as well as Bergmann 2008 An Introduction to Many-Valued and Fuzzy Logic: Semantics, Algebras, and Derivation Systems

Subjectivity may be involved in human choices in exact sciences, however, the subject matter itself is not subjective.

In any natural language there are ambiguous sentences. Let S stand for an ambiguous sentence and consider the statement:

"P=The sentence S means something"

I think that P cannot be either true or false. The truth-value for P must be greater than 0 and smaller than 1. In addition, its fuzziness arises from its structure and cannot be regarded as a subjective assignation. However, if you assign a concrete truth-value, for instance 0.5, then this asignation can be arbitrary and subjective. By contrast, the assumption of lying in the interior of the unit interval is objective, because of its fuzzy meaning.

Subjectivity, I believe, is associated with the reality of science, not fuzziness, although, in order to approach harmony and unity, we attempt to come to terms with every human being of our kind, our level, our interest. Every activity associated with man has subjectivity. Physics, biology, mathematics researches are proposed, conducted and reported by man. No two quantitative or qualitative research with similar topic and, methodology will come out with exactly the same findings, interpretations and recommendations, especially sentence construction and interpretation. Take for instance the history of civilization. It is not the same as the socialists, the capitalists, the Catholics, the Muslims, the poor, the wealthy, the quantitative researchers, the qualitative researchers, etc, view and accept it.. Where man is, subjectivity will be part of science, or research.

@Eddie, man utilizes subjectivity in the very nature of his thought process.

Precisely, Mike.

Now for the question of fuzziness relative to the nature of thought. One can argue that each choice we make, is made with some degree of certainty.

We can define a membership function m that maps the set of choices we make to the interval [0, 1]. And observe that all choices that are subjective are also fuzzy.

@ James Peters : This touches on the profound question whether (degrees of) fuzziness (are) is the same as (degrees of) probability. To my best knowledge this question is not yet explicitly, definitely answered (after almost 50 years of FST), although comparing the axiomatic frameworks of SP (subjective probability) with those of FL (fuzzy logic) I would say: completely different on several levels.

In order to get rid of the feeling of uneasiness which has come to be associated with the term 'subjective' , I always translate it (tacitly) into 'personal' ... and then all is fine.

Dear Dr. Vossen,

Thank you for your comment that this 'profound question' has not yet been answered 'explicitly'. In fact, this is why I have posted this question. As you have said, even after nearly half a century since 1965, the confusion is still there.

Perhaps objectivity is just the result of a self controled subjectivity of "personal" and collective effort in activities oriented to some final end. If subjectivity is not controlled the objective -or end- may not be obtained, or is simply distorted; if subjectivity is not applied then there is no action neither proiect, neither port to arrive, neither ethical limits. Many times the process conduce to unexpected fields and new interests. I agree with Paul Vossen words about personal, so please take his words personally: all this deals with responsible-subjective researchers attitude about their activity and the set of values that guide their practices and sense of life. Thanks, emilio

Paul,

You posts are great! You also have an interesting suggestion: translating "subjective" to "personal".

But Dear Professor James, subjectivity is definitely a personal matter; there is nothing fuzzy about this translation!

Dear Professor Jacic,

Why only in 'inexact' sciences, even in 'exact' sciences measurements may be subjective. The point is: what are those situations? Can we perhaps enumerate them?

We all have quite different personal as well as professional histories and thus come to this question with different philosophical presuppositions or what psychologists would call 'cognitive structures' and artificial experts call ''belief structures' or 'knowledge bases'. It is not always easy to figure out where exactly (inexactly?) we differ and, knowing that, come to a conclusion.

Now again to the term 'subjectivity'. I proposed to interprete it - if appropriate and intended - by the phrase 'dependence upon a subject' where subject is just another term for person or what is called agent by modern Artificial Intelligence.

Thus we avoid equating subjective with random, unpredictable, inexact, and possibly other concepts with which it could be confused.

Of course, it may turn out, that a subjective statement or fact or measurement may at the same time be random, or at least almost unpredictable or in any case inexact, but that is another claim, which should then be stated clearly and in addition to the claim that it is dependent upon an agent, a person, who told you the statement, the fact, the measurement.

What then is objective, objectivity? In my view it is the very special case that it turns out, that all credible agents, all persons whom you ask tell you the same 'story' (= statement, fact, measurement, ...).

Credibility of agents or persons is of course an important qualification, because you won't rely on agents or persons who might be lying or - worse - telling you different stories on different times (see the Hardest Logical Puzzle Ever of Smullyan/Boolos, a real brainteaser! http://www.thebigquestions.com/boolos.pdf).

How do we achieve such objectivity in both exact and inexact sciences? By constructing theories, methods, techniques and tools which do the observation and processing of real data for us! We call it standards and measurement devices (for the sake of simplicity, I include such measurements which give you true or false on statements, i.e. logic). Then in principle you haven't to ask another agent or person anymore about anything, but use those measurement devices. If all involved agents and persons do the same, you have perfect objectivity.

Achieving objectivity in this sense has not been an easy matter in all disciplines. Physical sciences have been lucky to deal with a lot of phenomena which are easy to measure. But you have to have the correct underlying theories, and we know that searching for the correct theories is still going on. Social and behavioral sciences started up much later, at the end of the 19th century, but were able to catch up very well (see the Handbooks of Mathematical Biology, Mathematical Psychology, Mathematical Sociology, etc.).

Why then is there still so much subjectivity, inexactness, randomness, unpredictability, etc.? I guess the answer is relatively simple: we don't get objectivity for nothing, we have to work hard in order to establish it, and we are just approaching it more and more in the long run. It is not that we can just decide to be objective from now on. No, we need to build and test the required theories, methods, techniques and tools in the first place. That may take another 1000 years or so (could be less, if mankind wouldn't spoil so much time and money on warfare and other counter-progressive activities).

Excellent summary and final statement. I fear that in education, too often subjective is identified as 'bad' and objective as 'good'. Paul has clarified the issue and removed those misunderstandings. Both can be bad or good depending on the quality of performance and the intent of the measure. Another misunderstanding in my circles is that numerical measures are always objective and therefore good and verbal descriptors are subjective and therefore bad. Any comments?

Dear fellows, I humble wolud like to present something out of the Gaussian curve, because I'm from an "inexact" science - health, dental surgeon. Only for a little more complicated logic or no logic: Decades ago a college professor of medicine from my campus (USP - Ribeirão Preto), in the field and area of nutrology, when asked in his presentation of the dissertation, where his research dealt with the consumption of poultry per capita by population in Ribeirão Preto, distributed by neighborhoods, including the periphery, one of the examiners, at the top of his knowledge wove harsh criticism logic to work because there was no statistical planning inside the study, which would make the work of Professor Dutra. Upon the candidate replied: " Well, if I say that the average consumption of chicken in Ribeirão Preto is 3 chickens per family, I was being flippant and inaccurate, since the intention of the research is to have social impact. Thus, while in the center of Ribeirão Preto all families who eat chicken every day (and other meat), there are neighborhoods where families (larger ones) those feed on one (01) chicken each two weeks or more. We're talking about nutrition, and while we have healthy individuals (or even obese) in the richest region of the city, some neighborhoods have many cases of malnutrition. For these people the application of statistical tests would be inhumane and only hinder the city administration." It is just a comment from someone who belongs to a large area that is not the exact sciences ... a happy and fruitful new year to you and your Families. And a lot of meat or good meal for all the world.

@Hemanta Baruah: ...subjectivity is definitely a personal matter; there is nothing fuzzy about this translation!

Perhaps you will agree that when choices are made on a personal level, we can define a membership function. Let X be the set of personal choices and associate the membership function m with each choice:

m: X --> [0,1]

Let x be a choice in X. Then m(x) represents the degree of certainty for the choice.

@ Mark Gould : "... misunderstanding ... that numerical measures are always objective and therefore good and verbal descriptors are subjective and therefore bad."

Indeed an unfortunate but pervasive misunderstanding. Goes back to how we learn numbers in our infancy and at school. Actually we start with one special number system: the natural numbers, for which an ordering, addition and multiplication and restricted forms of subtraction and division are defined. Later in secondary school some of us learn about negative numbers and fractional (or rational) numbers, so that all differences and quotients of such numbers can be built. That's all for 99% if not more of the population.

The point is, that what we learn is a formal system of numbers and how to manipulate them (

I perfectly agree with Ludwig Combrinck!

@Paul Vossen : "Example: Suppose that you find out that indeed 'zero' has a special status (the lowest possible value of a feature)..."

Actually, you want to take the plunge, and consider negative numbers tending toward minus infinity, instead of thinking of zero as the lowest number.

Many a new thing is coming up. This discussion will perhaps remove many a doubt about subjectivity.

@ James Peters : No, I don't plunge! The simple reason being, that I am imagining certain real-world phenomena (not numbers, yet) for which I can point out an absolute lowest bound, an absolute greatest bound and all sorts of levels in between, but no addition, no subtraction, or whatsoever. That is perfectly represented by the unit interval between 0 and 1.

I don't need infinity, which would imply that there is actually no real lower or upper bound. But I know, by assumption in this case, that there are hard lower and upper bounds!

Again, the fallacy here is to invoke and accept prematurely a given number system, and then asking what it stands for, instead of the other way around: investigating the properties of the features of your objects, and only then finding out, which kind of number system represents the feature space at best.

Paul,

Aha! Your post is excellent and raises lots of interesting, worthwhile points. When I wrote "take the plunge", I was thinking in terms of a novel approach to thinking about personal decision-making.

Perhaps you will agree that personal decision-making can be viewed in the context of possibilities. And the range of possibilities has an unbounded character, even though actual personal decisions (tied to actions) are very limited.

It is possible to find infinities in common (real-world) objects, not just in the cardinality of the natural numbers. For example, we know that the area of any circle you choose to draw is pi times the radius of the circle square. pi is a transcendental number.

James; I certainly agree with your view of personal decision-making: bounded for pragmatic reasons, though not necessarily in principle.

The other question is more subtle and may end up in a philosophical debate about Platonism, ontology and epistemology ;-)

Plato believed, as is well-known, that there exists such a thing as an ideal triangle, which presumably would have continuous (=real-valued) sides. All the observed triangles however are just "imperfect shadows" of this abstract triangle, and when you measure its sides you will always be bounded by your coarse, finite measuring devices, including your eyes.

Now what is reality? Is it the perfect triangle or the imperfect perceptions of it that impinge on our senses and are interpreted by our brains? It is equally well possible to explain the same relationship the other way around: what we see and believe to exist are the concrete triangles of finite granularity (according to e.g. quantum theory), but we have found a clever way to represent each one by a continuous mathematical object (an asymptotic object when granularity goes to infinte).

Assuming that the concrete triangle itself is continuous is confusing reality with an idealized model of it, according to non-Platonists. Thus it may be that our models of reality have properties which do not faithfully represent reality: they are overshooting actual reality, even if they work fine for all practical purposes.

If you like, you may try out these arguments on another example, involving pi, as you suggested: think about the game of dart and about the possible positions the dart may land on, a circular area. Someone wants to measure the probability that the dart lands somewhere. Now what do you do? (A) You run 1000 test with 10 good players and collect a huge, but finite number of approximate numbers of finite granualrity. Or (B): You analyze the idealized situation on paper using principles of geometry and mechanics. Who will have the better 'model of reality' in the end?

Paul,

Great post! I am glad that we are on the same page. Also, I am very interested in the question you have introduced. I am just on my way out (to shovel snow) but I promise to write again a bit later.

Paul, you write: ...Now what is reality? and ….Thus it may be that our models of reality have properties which do not faithfully represent reality: they are overshooting actual reality, even if they work fine for all practical purposes.

First the question of reality. The notion of reality has its origin in Kant's defense of science and scientific reasoning against the sense scepticism of Locke, Berkeley and Hume. Kant decided not to bother about the physical world and the untrustworthiness of the senses. Instead, he postulated that space and time are a priori (within us and prior to experience). And we endow our perceptions of the physical world and the physical world itself with a spatial and temporal character. Before Kant, the question What is reality? was not asked. It is just the world we live in.

You mentioned Plato, who is important because Kant was the ultimate Platonist. For him and for Plato, we never really know the absolute forms of things, we just try to reach higher and higher, gradually approaching a perception of the ideal world, the noumental world of Kant. Our world, for Kant, was phenomenal (all appearances) and behine the phenomenal world you will find reality, the noumenon.

For your second observation about models of reality, I think the place to start is to obtain precision in our perception of the world. This can be achieved by deciding on the features of objects of interest to us (e.g., luminance for picture elements, colour for precious metals such as gold, temperature for those who dwell in the snow country). Then it is our task to define a probe function for each feature and use the probe functions to extract feature values from the objects of interest In each case, we will obtain feature vectors containing numbers that are feature values. Feature vectors give us a precise description of objects of interest in the physical world. In sum, precise descriptions of objects lend objectivity to our personal choices concerning our world.

I will try to analyse the concept of subjectivity from a mathematical point of view. In classical mathematics (Set Theory & classical Logic) everything seems to be Platonically objective.

However after e.g. Boolean-valued models and in general nonstandard mathematics, a concept of "internal" and "external" observer has to be introduced. For example in infinitesimal analysis, an "internal sees *R as an *-Archimedean field, whereas the external one sees it as a non Archimedean. Similarly for the Poincare model of non-Euclidean geometry: an external observer sees the radius of the circle as a finite segment whereas the internal one as an infinity segment.

These examples along many others, are examples against the "objectivity" of classical mathematics.

Furthermore the above models give an infinity of models for the same structure, and this plurality destroy the uniqueness of the models in mathematics.

Summing all these examples we see that they introduce a kind of "subjectivity" in choosing one out of these models.

This subjectivity was enforced by the introduction of fuzzy sets, Topos theoretic models etc. Finally we must add that this subjectivity goes with the

introduction of postmodernism and the shifting from the mathematics of nature to the mathematics of humans.

Currently, we are in the era of the mathematics of organized complexity. Subjectivity is perhaps inherent in it.

Yesterday when I was talking about Platonism I suggested at one point: "what we see and believe to exist are the concrete triangles of finite granularity (according to e.g. quantum theory), but we have found a clever way to represent each one by a continuous mathematical object (an asymptotic object when granularity goes to infinte)."

That was before I read and heard about a revolutionary mathematical theory of physical nature proposed by a group of physicists around Nima Arkani-Hamed (the new "Einstein" according to some commentators). This theory is completely geometrical, doesn't need space-time concepts for its definition anymore and radically simplifies our notion of physical nature. The end of relativity theory and quantum theory as we know it now....?

If this theory holds its promises ( I mean, superstring theory once also claimed to be the big solver of all outstanding problems ...), then perhaps we have to fundamentally revise our cherished ideas about nature, and Plato may have been right after all (rehabilitation after some 2300 years): ideal geometrical objects are in a deep sense more real than the imperfect shadows we observe....

Paul, you write: Plato may have been right after all (rehabilitation after some 2300 years): ideal geometrical objects are in a deep sense more real than the imperfect shadows we observe….

From what you have written, you appear to accept Kant's view of the world as phenomenal (appearances). The idea of "shadows we observe" fits in with the Kantian view of world.

I think this is a good place to consider Occam's razor (eliminate unnecessary entities). In what you have written, "shadows" are unnecessary entities (an interpretation of what we observe).

James, it seems you are right, we should now start to systematically but cautiously apply Occam's razor to a lot of other (provisional, partial) theories and models, not just of philosophy and physics, but also of contemporary psychology and neuroscience. It all depends upon each other, I dare to say, and we should thus be very careful in our choice of wording and metaphors (including Plato's cave!), as we might let 'visitors' in who we just kicked out through the backdoor (hope this metaphor is not self-defeating...).

So yes, let's reconsider the mind-body problem, dualism, consciousness, subjectivity (ah, here we are again!), human perception and emotion, etc. If it turns out, that we can completely correctly describe and explain nature without reference to such old familiar concepts as n-dimensional space and time (a Cartesian concept in origin, which worked fine for some centuries) and at the same time arrive at models which are dramatically simplified (at least for those of us who understand the geometry of positive Grassmannians, I frankly admit, I don't) - well then we have to ask, what other concepts and views that we usually take for granted can be happily disposed of and replaced by others which better fit with the new Weltbild?

If space and time are just emerging phenomena, i.e. you can derive them from the new theory, this will be hot stuff for those, who have always said that there *is* no mind, no consciousness, no inner world, no perception in the way many philosophers and psychologist think and write about it. It's all emergence: complex by-products of our interaction with the one and only world we live in.

After all, we are ourselves part of this world and following Occam's razor principle, we should then first and foremost do our utmost best to describe ourselves (anatomy, physiology, psychology, ...) in terms which correspond as closely as possible to the simplest theory of nature possible.

For instance, a book like 'Observer Mechanics: A Formal Theory of Perception' (Bennett & Hoffman & Prakash, 1989), following a nondualistic approach, would be a good starting point for interdisciplinary research towards unification of physical, neurophysiologcal and psychological models of human perception.

Perception as an emergent phenomenon? Why not? And once you have this, other concepts like memory and thinking can follow. We have so many well-founded theories and models here and there (e.g. Eric Kandel, memory, Nobel Prize 2000), but it seems to me that they are scattered around and loosely coupled.

It may be time for a first, modest unification along the lines of thinking in modern physics (this is a long-term program, though, not an already achieved goal!) We have now for the first time in the history of science a very concrete case which shows that extreme simplication of our beloved but oh so complex theories is possible. That is encouraging and challenging.

Beware: I am not advocating an absolute reductionism of some sort, I am just contemplating the hypothesis, that many of our familiar models in behavioral science might turn out to be just beautiful but partial renderings of a more fundamental sort of model. If so, this would simplify and unify a lot of philosophical discussions, including those about subjectivity versus objectivity, with which we are back to the initial question of this thread.

Each "understanding" is subjective. If the subject is not hard science like physics or mathematics it is the way of argumentation which depends of the subjectivity of the author.

Dear Wilfred,

'Understanding' anything is subjective! Yes, it has to be true. Thank you for this answer.

I have then an interesting puzzle for you in order to find out where for you, subjectively, subjectivity ends and objectivity starts. As Wilfried Musterle puts it succinctly: "it is the way of argumentation which depends of the subjectivity of the author". This formulation with reference to "author" clearly refers to human beings, I would say.

Now replace "author" with "agent" in the context of modern knowledge-based systems (formerly called artificial intelligence). An agent in KBS can be a pure software system running on a standard (personal) computer or it can be a kind of robot with special-purpose hardware and software, e.g. one of those we have send to the moon or mars. I assume that we are talking here about an "intelligent" KBS with the ability to learn from experience and thus to expand and restructure its knowledge base, including any inference rules needed for argumentation.

You will certainly agree with me, that what this agent/robot "understands" and how it "acts" depends completely upon the current factual and procedural knowledge in its knowledge base, and may thus be quite different from what another agent/robot understands and how it acts.

Therefore in order to be consistent in our language and thinking, we should also attribute 'subjectivity' to our new friend, the agent or robot. In other words: he (or she?) is for all that we can tell a subject like you and me.

Do you agree with this argument? If not, why not? If yes, what other interesting things follow from that?

Paul,

the crucial point is "we are talking here about an "intelligent" KBS". We just do not have such intelligent agents an we do not know how to build such KBS. If we would have ... I agree heartfully.

Arnold Trehub argued consciousness would be only the representation of objects in a subjective relation in a volumetric space (retinoid space). In this sense your vehicle would be conscious. I disagree with this idea. Life and consciousness means "sense making" of input data and not only representation of data. Sense is a higher kind of organisation of object data. We can find qualia and consciousness if input data relates to these higher organized structures. Here is a lot of work ...

Paul, you write: ...Perception as an emergent phenomenon? Why not? And once you have this, other concepts like memory and thinking can follow.

Without meddling with the psychology of perception, Poincare's results approach offers an attractive means of interpreting what we observe. In his 1895 paper La science et l'hypothese (later a book), Poincare suggested an approach to modelling what he called the physical continuum (PC) and, in the process, introduce what he called a representative space (later named a tolerance space). His approach is to define a relation on sensation sensitivity (a carry over from Fechner's psychophysics experiments). He consider visual, tactile, weight-lifting and motile spaces. For example, let x, y, z be weights weighing 10, 11, and 12 gm, respectively. Weights

x and y appear to be the same, so x and y are put in a class A. But the difference between y and z is imperceptible. Hence, y and z are put in a class B. In effect, A and B have a common weight, namely, weight y.

Poincare's method works well, for example, in separating grey level pixel intensities in a digital image into overlapping classes.

@ Wilfried Musterle : "We just do not have such intelligent agents an we do not know how to build such KBS". Is this

(a) an axiom

(b) a theorem

(c) a working hypothesis

?

No snow to shovel here (not this time of the year anyway) but here is my take on reality. I will be a bit naughty and start by saying reality is not a philosophical question, as you can only investigate reality using empirical means (experiments, measurements, observations, etc.) and this rules out modern philosophy....this part of 'old' philosophy (Kant, Plato etc.) being replaced by 'physics'. So, I start of by saying forget them, that type of philosophy will not help to find reality. The theory of knowledge, metaphysics etc. forget it... that won't get us to 'reality'. This sort of narrows things a bit having cast away philosophy. We end up with our perceptions and understanding of our observations. These are limited by our mental abilities, interpretations, scientific methodology and instrumentation capabilities. These, fortunately seem to be improving and expanding every year as the scientific community relentlessly investigates, probes and analyze 'reality' in order to measure it, categorize, digitize, discretize and all the other 'izes' whatever is observable and what is predicted to be observable with the odd serendipitous surprise in between.

So, present reality is basically our present human observations, scientific observations and understanding of our environment. I say basically, as part of our reality remains 'unscientific' and is based on other things (e.g. remnants of old philosophy). But no need to run to Kant for reality.

Dear Wilfred,

Not just understanding, thinking too is subjective. Right?

Ludwig,

Great post! Basically, I agree with your view of reality, namely, what we observe.

However, physics replacing philosophy as the observer (and caretaker) of reality is stretching things a bit.

You mention that Kant (and Plato) represents old philosophy. To come fuil circle on this relative to physics, Einstein reportedly kept a copy of Kant's Critique of Pure Reason under his pillow. Perhaps you can see why this bit of folklore about Einstein makes sense. As far as I know, the space-time model is imposed by us on the physical world as a means of viewing the world relativistically.

@Paul: If we are talking about agents we talk about artificial KBS. So it is a fact.

Only in this case we can understand the working principles.

Naturally there are some billions natural KBSs on the planet ...

@Hemata: Yes, all cognitive processes even perception.

Selection and design of meaduring devices in physics are likewise subjective. We have to interpret prior understandings in terms of the current shared view of reality and design measurements to suit. So there is a subjective element there. It is possible (though not easy) to imagine a trajectory ofphysics theories that leads us to a different place in our description of reality. In the end it is about how successfully the theories predict outcomes 8f the next experiment rather than represent reality. The more esoteric our level of explanation of reality becomes, the more subjectivity intrudes cf string theory.

PS we see occasional researchers step outside the norm of current thinking eg Einstein wrt Newton and then Bohr et al wrt Einstein and these, to me, represent the uncertainty in our current view of reality. Not that I believe in a purely philosophical view of reality, but that there may be many ways to represent it that works witinh the parameters we choose (choice is subjectivity). Probably, the number of representations will decrease as we move forward??? Some of this comes from Popper.

Hi James, I thought you would see it as stretched!

However, we must keep in mind that philosophy has changed and so has science. Reality is no longer the realm of mind experiments only (what did Kant measure?) and I am leaning more to what a blind person would do to find reality, touch, listen, smell, interpret and think afterwards, then start all over until I am happy with my reality.

Spacetime is of course just a current interpretation of the physical world as you say, all our models of the physical world and universe create our 'reality'. Einstein kept a copy of Kant's Critique of Pure Reason under his pillow but downplayed the influence of Kant on his work. It probably did have an influence, as he liked Kant's writings. I read Kant in my teens and found it fascinating, I suppose everyone would, also Einstein. It does not change reality for me however, that is more personal. Every individual experiences his own reality. I suppose there is a collective reality as well.

@ Ludwig Combrinck. Hm, interesting, you wrote: "I suppose there is a collective reality as well." Suppose by way of thought experiment, that there is no such thing as a collective reality. How would that change your individual reality?

May be we should philosophize a bit more ... "be or not to be: that is the question" Prince Hamlet, William Shakespeare... or is freedom an alternative?

Dear Professor Wechsler,

In other words, you are saying that subjectivity is associated with every field of knowledge.

Paul...for a bit of a gedanken experiment, I am not sure one can dislocate individual realism from the collective, that would be an anti-Borg attempt, and probably disallowed. If you have no collective, i.e. you are really on your own, independent (and no one on Earth is that) you could probably truthfully have no collective reality. In our case however, our knowledge is a collective knowledge...built up by the collective through the ages.

@ Ludwig,

religion is such a collective "reality".

@ Wilfried, yes, one could say that. Religion is of course a bit beyond reality as it delves into the immeasurable to some extent. It does not mean that the beyond part is not real, it is just not measurable (on an instrumental, scientific level).

@ Ludwig Combrinck. Hm. If our knowledge is a collective one, built up by the collective through the ages, as you say, how comes, that so many people disagree with each other? Or is disagreement itself part of the collective understanding? Wouldn't that in the end - when one really thinks through - lead to a collective belief that accepts everything, i.e. doesn't distinghuish between truth and falsehood, because from Belief A and Belief non-A follows Falsehood and from Falsehood you may infer anything you like (by a simple logic theorem)?