…It's in words that the magic is—Abracadabra, Open Sesame, and the rest—but the magic words in one story aren't magical in the next. The real magic is to understand which words work, and when, and for what; the trick is to learn the trick.

…And those words are made from the letters of our alphabet: a couple-dozen squiggles we can draw with the pen. This is the key! And the treasure, too, if we can only get our hands on it! It's as if—as if the key to the treasure is the treasure!

John BarthChimera

[1] The same idea is pervasive throughout all of engineering. For example, electrical engineers use many different languages for describing circuits. Two of these are the language of electrical networks and the language of electrical systems. The network language emphasizes the physical modeling of devices in terms of discrete electrical elements. The primitive objects of the network language are primitive electrical components such as resistors, capacitors, inductors, and transistors, which are characterized in terms of physical variables called voltage and current. When describing circuits in the network language, the engineer is concerned with the physical characteristics of a design. In contrast, the primitive objects of the system language are signal-processing modules such as filters and amplifiers. Only the functional behavior of the modules is relevant, and signals are manipulated without concern for their physical realization as voltages and currents. The system language is erected on the network language, in the sense that the elements of signal-processing systems are constructed from electrical networks. Here, however, the concerns are with the large-scale organization of electrical devices to solve a given application problem; the physical feasibility of the parts is assumed. This layered collection of languages is another example of the stratified design technique illustrated by the picture language of section 2.2.4.
[2] The most important features that our evaluator leaves out are mechanisms for handling errors and supporting debugging. For a more extensive discussion of evaluators, see Friedman, Wand, and Haynes 1992, which gives an exposition of programming languages that proceeds via a sequence of evaluators written in the Scheme dialect of Lisp.
4 Metalinguistic Abstraction