The Content Embedding family of patterns was originated by means of a retrospective view of our previous experience in creating educational software and games. The two main experiences from which this patterns have been derived are:

1. The design, development and experiment of a Computer Algebra System called L'Algebrista (see Michele Cerulli's PhD thesis for a full acount of this issue )
   
2. The design, development and experiment of LEGO RCX robots and of Toon Talk computational objects, for the Weblabs project .

In the first case, we had to develop a software which could be used to introduce pupils (this were our Educational objectives) to algebra as a theory, thus to the ideas of theory, axioms, theorems and proofs. In particular we wanted to vehiculate the idea that an axiom and a theorem, are both, alements of a theory, and instruments for proving new theorems. All these concepts are "abstract" and the challenge was to make them tangible/objectified/usable in order to make them concrete for pupils, and we tried to do this by embedding them in a software, L'Algebrista. In order to do that we:

• individuated what aspects of each single concept we wanted to be embedded in the software and thus to be learnable by the pupils by means of activities involving the software
• defined how to represent such aspect by means of the software
  
• for each such aspects, we defined how they should have been experienced directly with the software, in other words, we defined the actions to be performed with (or on) the software that could make esplicit, or involve directly, the concepts we were interested in.
• for each of such aspects we defined what activities could be set up (with or without the software) in order to make the embedded knowledge correctly learned by the pupils.

In terms of typologies, this can be described as:

• indviduated a Mathematical Content to be embedded in the software 
• defined what Representations of the chosen mathematical content should have been implemented in the software
  
• defined the relationships between such mathematical content and the software (relationship and How Embedded)  
  
• defined a set of Modalities of employment for exploiting the software to achieve the chosen Educational objectives
  

A review of this process led us to define the Content Embedding pattern.

In the case of the Weblabs project we proceeded in a similar way, but with different Educational objectives and aiming at the design of games, not of softwares in general (in fact we did it also with LEGO robots).
In the particular case of the LEGO robots, we proceeded similarly to the case described above, but we focused particularly on the relationship and How Embedded aspects, designing robots with different levels or transparency with respect to the embedded Mathematical Content, which concerned probability. In fact we had a robot with a random generator controlled by the user: in this case, concept of randomness was  embedded in actions, and embedded in toy and the relationship between the user, the toy, and the concept of randomness was considered to be explicit. Another robot had an authonomous source of randomness, so the user could only observe it, but such a random generator was external to the robot, and its functionaing was clearly visible for the user who was asked to discuss the random nature of the robot: in this case the concept of randomness was embedded in toy and embedded in focus and such relationshipt was explicit. The third robot didn't have any visible random generator, in fact it used a random function of its RCX computer, and pupils were required to discuss its behaviour: in this case the concept of randomness was embedded in toy in an implicit way, and embedded in focus in an explicit way.

In a similar way, we embedded probability related concepts in a set of computational objects, and drawing from all these experiences we derived the Content embedding in toy pattern. However, still in the Weblabs project we developed the Guess my Garden game, deriving it from the Guess my Robot robot game. In the transition from one game to the other, we realized that in the Guess my Robot game, some of the addressed mathematical contents were embedded in its rules and could be learnt by means of discussion of such rules. Thus we decided to keep this kind of relationship and to have some aspects of our mathematical content (probability) embedded in rules in an implicit way, and embedded in focus in an explicit way. A revision of such design process led us to define the Content Embedding in Rules pattern.