Why is there something like ‘meaning’? Because there is something like encoding, there is something like meaning (the decoding). Now, language is an encoding. So there is something like meaning, in, or rather for, language.
I have to have a representation of ‘tomorrow’ to think about tomorrow or to think about something involving tomorrow. It could be a word, it could be a visual, it could be an action…..So (mental) data itself is an encoding (in the form of the representation). Sensory data…..real data ….is that encoding? No, since a tree doesn’t stand for a tree; it is a tree. There is a real tree and there is a tree in thought. There is a real tomorrow and there is a tomorrow in thought. So thoughts are an encoding of reality. And systems which do that are minds. So a phone which can take a snap of a tree is a mind in some sense since it can encode and store reality (the real tree). Language is a representation and writing media are minds. Language is the software of reality, which can exist on almost anything – rocks, water, brains, computer hard disks etc.
So thought and language are 2 representations (like softwares) operating in conjunction, in the human mind.
Any representation needs rules. The representation that thought is, has natural rules – for the visual, auditory etc. correspondences to reality which are stored in temporary/long-term memory, for the purpose of manipulating. The representation that language is (which is used for communication) has man-made rules.
So in my mind I have 2 representation-systems of reality – thought (natural-rules) and language (learned man-made rules). Thoughts are enabled by my sensory perceptions of reality. (Manipulation of the same requires memory). Language is enabled by the abstract, learned rule-set in my memory. This rule-set is again composed of knowledge, which is again a natural representation of reality and hence like the thought-system of representation. This last sentence just shows, as we know, that language is a sub-system of thought.
How are representations manipulated? Suppose I have a representation (natural) of a certain reality, say, a cow, in my mind (so its a thought). And I have another natural representation (hence a thought) of a reality – say, a man. Sheer physical superimposition/closeness (spatial-proximitizing) can make me combine the representations of the cow and the man. In other words, ‘bringing together’ / ‘combining’. Spatial alterations – in wholes / parts – brings about legs being bent (in this case, in parts – just legs). One of the possible variants of the total of such few alterations is a man sitting on a cow. Hence my “new” thought.
Hence the operators of manipulation of representations are things like – rotate/cut/crop/magnify/twist/…etc. in parts and wholes; and combine/super-impose/add/….etc. (which we have on computers today!) This is thinking. The similar applies for manipulation of language-representations, though that would be more rigid because of a heavy rule-set library (that too man-made and acquired) and lesser inherent flexibility of the representations (words), as compared to the natural-rules’ representations (e.g. visual) like those of the thought-system.