What is the main difference between "information" and "theory?"

This is an essential question in many ways.  I think the questioner has some good insights into the problem area, but the questions need sharpening.  I hope the questioner won't take it  personally if I start by examining the question  and its assumptions, because they pretty well represent the  confusion inherent in the current "standard notion" of information.  To start with, I'd recommend backing up and acknowledging the confusion and ambiguity is both warranted and pervasive.  There are literally dozens of different and incompatible definitions (theories, if you don't mind) for "information" and "theory".  The Shannon/Weaver theory is essentially little more than a technical analysis of data transmission and coding.  Data can be transmitted, provided there is an already-agreed-upon convention to construct the machines that encode and decode that data.  The Shannon model forgets about the code machines and other shared world elements of the sender and receiver.  And as we know, those -- the context -- are the essential aspects of knowledge and meaning, at the "human scale". Information is an effect, a process, not something that can sent and received -- if someone says it can, they are using the word to refer to "data".  One of the best surveys of various models (about "information" as well as "theory") is in the book http://books.google.com/books?hl=en&lr=&id=XYX_RV7Wy9QC&oi=fnd&pg=45#v=snippet&q=%22concepts%20relevant%20to%20information%20behavior%22&f=false, by Donald Case. So the question as framed has ambiguous premises.  On the one hand, it starts with a machine-like (or rather, computer-like) definition of information.  More specifically, the definition implies a von-Neuman type computing model in which computation is separated from data:  the ghost of the Turing Machine, where information is what can be coded on the tape;  there's separate processing unit that itself has no embedded data contents and is fully "generic". Then there's a huge jump (or disconnect) to the human scale questions regarding knowledge, being informed, acting socially, and so forth.  The Turing machine computer theory and its adjacent theories of information suddenly finds itself vastly underpowered (or over-simplified) to model what goes on in that domain. Assuming that machine computation and coding models apply to real-world knowledge and information behavior tries to connect two dots that are just too far apart.  One could attempt an argument by principle, but this is the Information Age equivalent of Archimede's triumphant insight about the power of the lever:  "Give me a lever and a place to stand, and I will move the world!" Sorry if this doesn't "resolve the ambiguity", but it does help frame the question. As it stands, the machine logic and human scale definitions of information and theory cannot be directly bridged.  How about indirectly?  I think that's where things get interesting. Followup Questions If we transmit "information" as if it were data, where is the "codebook" that defines its meaning and interpretation? Can we treat theories as data?  Can we represent theories explicitly?  Is that needed in order to communicate them? What else does a system need to have for the phenomenon of "information" to be present in it? What is the http://scholar.google.com/scholar?q=%22extended+mind+hypothesis%22&btnG=&hl=en&as_sdt=0%2C48? Can humans get better at meta-cognition, and if so, how?

Just a simple way of looking at it, 1) From the perspective of greater-lesser Set of Information > Set of Theories (all theories are based out of some information), but not vice versa. 2) From the perspective of sequence Information comes right after "observation" & is immediate before we "form" Theories about them, So, Reality --> Observation --> Information --> Theory --> perception , so on & so forth