The difficulty in understanding Wittgenstein ('W') is threefold. First, the easiest way to understand him these days is to start at end of his thought and work backwards. Second, W himself had very little to say on metaphysics and theories of meaning, which has led to totally contrary consequences by other thinkers building on his ideas. And third, which is the point people usually start at, is that he started with a very different conception, then changed his mind half way through his life and recast his earlier thought in a new light.

One of the last things W wrote about was color. Realists just think of color as a sense experience, which is physically described in terms of certain electromagnetic wavelengths, and any dispute is resolved by refining the boundary conditions on allowable mixes of wavebands. On the other hand, the significance of color is social too. For example, in the USA red frequently has the connotation of danger or 'stop.' In China, red is the color of a party or parade, and therefore cheerful instead. But W instead took a middle position on it, and instead concentrated on how two people reach an agreement on what the color red might be. With regards to whether objects in the world are red or not, he concluded, from the nature of how agreements are reached, the true nature of reality can only be known 'mystically.' MANY people have a problem with that, so the next step is to consider how W reached that conclusion.

During W's life, the most dominant view of language was that it is 'descriptive.' An 'apple,' for example, is known to us by a description of a real-world object that exists in our minds. So when we want to know if it is true that there is an apple is red, we 'look up' the descriptions of 'apple' and 'red' in our minds, then we look at the real world to find the apple and see if its color corresponds with our mental description of red. This approach, popularized by Russell and Whitehead, was immensely successful in creating new approaches to representing language symbolically, and so this 'theory of descriptions' became rather accepted as the way language correlates with reality.

W really did not like the theory of descriptions at all, because to him, the resulting symbolic language of propositional logic creates an artificial and unreal perspective on the nature of 'natural language.' W instead viewed language as a 'game' or 'tool' whereby people can communicate their intents, and cause changes in the world, in ways that can be totally unrelated to dictionary-style descriptions of words.

W's alternative was very difficult for me to understand until I learned, of all things, a security protocol used in computer software called the 'Diffie–Hellman key exchange' for inter-user authentication. By this method, two people can have 'private color codes' to encode data for exchange, and yet they each can understand each other without knowing the other person's private color code. To do so, they both mix their private colors with a public color; and the tones of the private and public colors are chosen so that, when they each receive a colored message from the other, they can each subtract their private color from the message color, extracting the same color that the other sent. The result is a 'shared secret communication' where both have the same color, but don't know the secret color of the other; and that remains the initialization of all secure connections on the Internet today (for the full mathematical explanation, see https://en.wikipedia.org/wiki/Diffie%E2%80%93Hellman_key_exchange#Cryptographic_explanation).

The idea of 'shared secrets' helped me understand W as follows. In this example, Alice and Bob both have a secret intent that the other does not know, but they can still communicate effectively:

* Alice [secret] is hungry ->
-- Alice says 'Bob please bring in groceries from car.'
* Bob [secret] hears 'Alice wants me to get dressed to go outside.' ->
-- Bob says 'It is raining'
* Alice [secret] hears 'Bob is not hungry.' ->
-- Alice says 'it is not raining.'
* Bob [secret] hears 'Alice insists I get dressed to go outside.' ->
-- Bob says 'if you think it's not raining, you bring in the groceries, and in exchange I'll start cooking dinner.'

Through W's method the outcome is happy. Yet had Bob instead been a realist, a long argument would have ensued as to the nature of the meaning of 'rain,' with disputes about the difference between misty condensation, rainfall, and sleet. So the realist, following Russell's descriptivist theory, believes the solution to the contradiction is to converge on a common definition that "rain is defined as water drops, of size greater than XX mm3, falling from the sky at a rate of YY drops per square meter per second, above the temperature 0 Celsius." Such debate might have kept Bob from getting dressed, but meanwhile, Alice would be increasingly hungry, which is one explanation why realists can experience troubled marriages.

Wittgenstein's point is now clear:
* It is totally irrelevant to the conclusion whether the proposition 'it is raining' is true or not.
* It is certainly irrelevant what the correct description of rain might be.
* Bob is not told Alice is hungry.
* Alice is not told Bob doesn't want to get dressed.
And yet even so, Bob's 'positivist' conclusion is logically coherent, the result is happy, and there has been effective communication.

It's impossible to explain this in terms of the 'descriptive theories of reality' that materialists frequently attempt to apply to all aspects of the amazing world in which we live. Yet strangely, some theorists have tried to warp W's idea of 'positivism' into a disproof of the existence of meaning or a dualist mind/body reality. In fact, W has, as I started by saying, no intent to state 'any theory of meaning' or 'metaphysical grounds' at all, because, as he tries to point out, that denies the ability to understand the real nature of the conversation between Alice and Bob.

That is to say, W holds that a theory of meaning and metaphysics is NOT NECESSARY for effective communication. Just because they are not necessary does not imply that other forms of communication could be based on theories of meaning. Rather, Wittgenstein has instead expanded the scope of that which can be understood in language beyond that which theories of logical meaning and metaphysics can explain.

Therefore, he concludes, if one attempts to make a single, general, total statement about ALL communication, the most one can do is say that language exists, and the totality of 'the world' as far we know it for sure, is only the facts we state in language. About all other things about which we do not speak, we cannot really say what is true or not. In his earlier narrower theory of logical positivism, he held that statements about 'intrinsic states' (such as emotions) fall outside the boundaries of that which we can speak and know to be true; but his later thought, by adding ideas of language games and linguistic tools; increased the domain of knowledge beyond that empirically verifiable in some largely undefinable way.   

There, then, is a return to the beginning of W, and where I end...Except to say, maybe Bob actually knew Alice was hungry. And maybe Alice knew Bob didn't really want to get dressed. Whatever the metaphysics and nature of meaning, however, they are both still wiser to pass over that in silence. :)