Out of an infinite number of potential differences, how do we choose which differences will make a difference? In other words, how do we decide which information is relevant?

It is context that defines relevance, and goals that define context. So it is our goals - the things we aim at - that confer sense, and that define the things that we see. Inasmuch as a 'context' is a kind of story, then it follows that we see what our stories allow us to see.

The realist believes that an object defines itself - in other words, it is a select number of inherent properties that make the object what it is, and it is these properties that we perceive. The object is not conditioned by the subject.

The pragmatist believes that the subject defines the object: that our goals - and the contexts that are conditioned by these goals - affect which information becomes relevant. The object is entirely conditioned by the subject.

There are an infinite number of ways to perceive or construe a given situation, and an infinite number of potential consequences of a given action or event […] Every “object” can be classified, even perceived, in an infinite number of manners. Every object must therefore be regarded as something infinitely complex – objectively, intrinsically.

We live in a virtual sea of complexity and uncertainty.

[…] how can a given entity be segregated from the parts that compose it, the situation of which it is a part, and the other entities to which it is related?

It has become painfully clear that whatever constitutes the definitive boundary of the objects that we manipulate both concretely and abstractly is not something simply intrinsic to those objects.
Events are simple and distinct only insofar as their relevant features are framed, a priori, by the constraints of an operative context.

We frame an object by cutting away vast swathes of information, indisputably associated with that object, but irrelevant to our current purposes, however those might be subjectively construed. Lacking subjectively determined constraint, therefore, even the most powerful information processors cannot compute the explosive complexity of the object – cannot even reliably define the edges of a regular solid, say, under changing conditions of lighting or position.

This means, to put it bluntly, that the concept of “object” itself is somewhat illusory. The world does not present itself neatly packaged into pre-existent categories, available for direct human perception and manipulation.

Nothing at all can be understood in the absence of a structured and subjective frame of reference. 

[Jordan B. Peterson]
‘Complexity Management Theory: Motivation for Ideological Rigidity and Social Conflict’, in Cortex, December 2002, p. 432-3

Sokolov [...] noted that creatures exposed to novel or anomalous stimuli responded with eye movement, or alterations in galvanic skin response, or “depression of brain-wave rhythms,” [known as the ‘orienting response’] and believed that these alterations were not due so much to “incoming excitation” as to signals of discrepancy which develop “when afferent signals are compared with the trace formed in the nervous system by an earlier signal.”

It is difficult to determine how an organism might manage anything as complex as an orienting response – which, as Sokolov described, might be elicited by “the slightest possible change” in a given stimulus – without first constructing an elaborated and detailed model of the world.

A given entity or object is actually very difficult to “perceive” – not so much because things in themselves lack structure, as classical nominalists might have it, but because that structure is so rich and variegated that it may be endlessly and variously construed.

Although there appear to be certain “basic level” categories that “leap out at us, and cry out to be named,” in the developmental psycholinguist Roger Brown’s terminology – so we naturally apprehend the table, instead of each of its four legs and its single flat surface – we do not know precisely how our perceptual and cognitive systems manage this apprehension.

[...] two things differ and are the same in as many ways as there are potential things to which they might be compared: books in a library, for example, might be categorized by the total number of “e’s” they contain, or by their age, or thickness, or by the number of atoms of selenium on the first page of their preface, or by how closely they approximate the weight of the average book.

It might well be objected: such classificatory strategies are ridiculous – but the problem with that objection is that the lack of utility of a given strategy is a judgement of value, not a necessity drawn by logic or by any conceivable objective standard (as any “objective” judgement requires the a priori establishment of value-based criteria to judge by).

And, if it is judgement of value that determines the validity of classification (or even of perception), then it could easily be that functional utility determines the nature of the object. 

A chair is not a chair because it shares a set or even a subset of identifiable objective properties. A chair is a chair because a person can sit on it. This makes a chair a tool, – a tool useful for specifically human purposes – and not a thing.

In the absence of a specific goal, or at least the possibility of a specific goal (which means in the absence of arbitrary and value-predicated constraint) the universe does not reveal itself as structured, or it reveals itself as too complexly structured, which is very much the same thing.

“Objects” are therefore not the simple constituent elements of the objective world, directly and simply perceived, but tools apprehended with difficulty, in the service of specific goals (and, furthermore, tools that may be perceived at very different levels of resolution). 

[Jordan B. Peterson]
‘Complexity Management Theory: Motivation for Ideological Rigidity and Social Conflict’, in Cortex, December 2002, p. 434-5

[Jung] points out that there are two worlds. We might call them two worlds of explanation. He names them the pleroma and the creatura, these being Gnostic terms.

The pleroma is the world in which events are caused by forces and impacts and in which there are no "distinctions." Or, as I would say, no "differences."

In the creatura, effects are brought about precisely by difference.

In fact, this is the same old dichotomy between mind and substance.

We can study and describe the pleroma, but always the distinctions which we draw are attributed by us to the pleroma. The pleroma knows nothing of difference and distinction; it contains no "ideas" in the sense in which I am using the word.

When we study and describe the creatura, we must correctly identify those differences which are effective within it. The creatura is thus the world seen as mind, wherever such a view is appropriate.

And wherever this view is appropriate, there arises a species of complexity which is absent from pleromatic description: creatural description is always hierarchic.

[...] there are differences between differences. Every effective difference denotes a demarcation, a line of classification, and all classification is hierarchic.

In other words, differences are themselves to be differentiated and classified.

[Gregory Bateson]
Steps to an Ecology of Mind, p.461-3

The way we behave in any situation has a lot to do with the context. We whisper in hospitals and become anxious in police stations, sad in cemeteries, docile in schools, and jovial at parties. Contexts control our behaviour, and our mindsets determine how we interpret each context.

Many of the contexts that affect us most deeply are learned in childhood. For instance, our early visual exposure to the world may actually shape what we later see.

A controversial study of Euro-Canadians brought up in urban settings where buildings surround them with right angles, and with Cree Indians raised near tents and lodges that have many shapes and angles, suggested that the effects of early visual context may be lasting. In adulthood, right angles could be seen better than other line orientations by the Euro-Canadians. At the same time, they seemed to have less visual acuity for oblique orientations than the Cree. From the beginning, the Cree have a different mental landscape which may allow them to take in a greater variety of visual cues.

The location of context in our perception was vividly illustrated in an experiment conducted by psychologists David Holmes and B. Kent Houston. With the permission of the subjects, they administered mild electric shocks to a group of people, half of whom were told to think of electric shocks as new "physiological sensations." Those who thought of the shock in this way were less anxious and had lower pulse rates than those who were not given the prior instructions.

The same situation or stimulus called by a different name is a different stimulus. Roller coasters are fun but bumpy plane rides are not.

It has long been known that values create a context that influences sense perceptions. In 1948, Leo Postman, Jerome Bruner, and Eliot McGinnies used a machine called a tachistoscope to flash words very quickly on a screen. These words were associated with various values.

For example, subjects were shown political words such as govern, citizen, and politics; religious words such as prayer, sacred, and worship; and aesthetic words such as poetry, artist, and beauty. Words were shown to subjects in random order.

Despite the fact that the chosen words were equally familiar, the speed with which subjects recognized the words varied as a consequence of the subjects' values, as measured by the same Allport-Vernon Scale given earlier to subjects.

The higher a subject's score on a particular value, the more quickly he or she recognized the word. Politically oriented subjects, for example, recognized political words sooner than the artistically oriented. The context created by the subjects' values appeared to affect their visual ability.

[Ellen Langer]
Mindfulness, p.35, 36, 38, 39, 40

A researcher named Jules Davidoff traveled to Namibia [...] where he conducted an experiment with the Himba tribe, which speaks a language that has no word for blue or distinction between blue and green.

When shown a circle with 11 green squares and one blue, they could not pick out which one was different from the others — or those who could see a difference took much longer and made more mistakes than would make sense to us, who can clearly spot the blue square.
But the Himba have more words for types of green than we do in English.

[...] Davidoff says that without a word for a color, without a way of identifying it as different, it is much harder for us to notice what is unique about it — even though our eyes are physically seeing the blocks it in the same way.

[Kevin Loria]
What is blue and how do we see colour? 

A text is a place where the irreducible polysemy of symbols is in fact reduced because in a text symbols are anchored to their context. 

The medieval interpreters were right: one should look for the rules which allow a contextual disambiguation of the exaggerated fecundity of symbols. Modern sensitivity deals on the contrary with myths as if they were macro symbols and - while acknowledging the infinite polysemy of symbols - no longer recognizes the discipline that myths impose on the symbols they involve. 

Thus many modern theories are unable to recognize that symbols are paradigmatically open to infinite meanings but syntagmatically, that is, textually, open only to the indefinite, but by no means infinite, interpretations allowed by the context.

To recognize this principle does not mean to support the "repressive" idea that a text has a unique meaning, guaranteed by some interpretive authority. It means, on the contrary, that any act of interpretation is a dialectic between openness and form, initiative on the part of the interpreter and contextual pressure.

Medieval interpreters were wrong in taking the world as a univocal text; modern interpreters are wrong in taking every text as an unshaped world. 

Texts are the human way to reduce the world to a manageable format, open to an intersubjective interpretive discourse. Which means that, when symbols are inserted into a text, there is, perhaps, no way to decide which interpretation is the “good” one, but it is still possible to decide, on the basis of the context, which one is due, not to an effort of understanding “that” text, but rather to a hallucinatory response on the part of the addressee.

[Umberto Eco]
The Limits of Interpretation, p. 21

There is no bottom line. Our choice of terms depends on the purpose for which we want to analyse the world at the time.

[Mary Midgley]
Science and Poetry, p. 195

Complex systems are usually open systems, i.e. they interact with their environment.

As a matter of fact, it is often difficult to define the border of a complex system. Instead of being a characteristic of the system itself, the scope of the system is usually determined by the purpose of the description of the system, and is thus often influenced by the position of the observer. This process is called framing.

Closed systems are usually merely complicated.

[Paul Cilliers]
Complexity and Postmodernism, p.4

[...] in any given social situation it is frequently the case that many different transactions are taking place at one and the same time between members of the same group. Not one game is being played, but several, and [...] the problem about real life is that moving one's knight to QB 3 may always be replied to with a lob across the net.

Even when we can identify with some certainty what game is being played, there is another problem. In real life situations, unlike both games and the examples in books about game-theory, we often do not start with a determinate set of players and pieces or a determinate area in which the game is to take place.

There is - or perhaps used to be - on the market a cardboard and plastic version of the battle of Gettysburg which [...] had this peculiarity, that a moderately good player taking the Confederate side can win. Yet clearly no player of war games is likely to be as intelligent at generalship as Lee was, and he lost. Why?

The answer of course is that the player knows from the outset what Lee did not - what the time scale of the preliminary stages of the battle must be, precisely what units are going to get involved, what the limits to the terrain are on which the battle is to be fought. And all this entails that the game does not reproduce Lee's situation.

For Lee did not and could not know that it was the Battle of Gettysburg - an episode on which a determinate shape was conferred only retrospectively by its outcome - which was about to be fought.

Jeffrey S. Milstein and William Charles Mitchell (1968) constructed a simulation of the Vietnam War [...] What is striking [...] is their response to their second source of failure, one which they do note themselves: their predictions were radically upset by the Tet offensive. The response of Milstein and Mitchell is to speculate as to how future studies might be extended so that the factors which led to the Tet offensive might be included.

What they ignore is the necessarily open and indeterminate character of all situations as complex as the Vietnam war. There is at the outset no determinate, enumerable set of factors, the totality of which comprise the situation. To suppose otherwise is to confuse a retrospective standpoint with a prospective one.

To say this is not at all the same as saying that all computer simulation is valueless; but what simulation cannot evade are the systematic sources of unpredictability.

[Alasdair MacIntyre]
After Virtue, p.115-16

Related posts:-
Everything is Connected
The Eternal Ideas
Guiding Fiction 
Mind Your Language
Everything and Nothing 
A Difference that makes a Difference
Small Part, Large System