STATUTE Corporate, a sophisticated development tool as a result of ten years of research at the Australian National University and the University of Canberra.

"LISP" is an abbreviation for "LISt Processing." As its name implies, it works primarily on lists -- ordered sequences of elements. A list-element can be an irreducible item (called an "atom") or it can be a list. Lists are flexible structures. You don't have to specify their characteristics in advance. You can use them to represent all kinds of things, and you can use list-related operations to represent all kinds of processes.

I believe that these characteristics of lists (along with other language features of course, like recursion and an interactive development environment) makes LISP so appealing to AI developers. The bases of intelligent behavior are complex. The processes and structures we model, therefore, can present complications that we aren't always aware of when we begin modeling them. It helps to have a language that allows you to create on the fly.

And Prolog? Designed to work with symbolic logic (its name stands for "PROgramming in LOGic"), this language is widely used in Europe and Japan. Its popularity is on the rise in the U.S. Prolog is a declarative language. You don't tell it how to do its work (as is the case with a procedural language). Instead, you tell it what you want it to do. For traditional programmers, this can be a hard concept to grasp. For Prolog programmers, it represents liberation from procedural constraints.

Prolog's declarative nature (and the processes that support it) make it an ideal language for modeling complexity. For many kinds of applications, Prolog programmers can concentrate on high-level concepts and not concern themselves with nuts and bolts.

To demonstrate the range and power of these languages, our language-oriented articles span the spectrum from the theoretical to the practical. The Feature Article, "Logic Programming and Natural Language" introduces a series on a Prolog-based technique that's ideal for analyzing sentences (and other sequences of symbols that follow pre-defined rules). In "Prolog in the Real World," Al Roth and Clive Spenser tell us about a number of successful commercial applications of logic programming. In one of our two Vendor's Forums, Dennis Merritt details the thought process behind Amzi!'s development of an API that interfaces Prolog with C and C++. In the other Vendor's Forum, Tony Wei and Costas Tsatsoulis of CECASE discuss MEM-1, a case-based reasoning development environment based on LISP. Lisa Lewinson rounds out our language features with "LISP in the Process Industries." Lisa's article is a look at a commercially available LISP-based Windows system that helps planners and schedulers in a variety of industrial contexts.

We include other AI areas in this issue as well. Neural net aficionados will enjoy Anthony Kempka's "The Neural Net Connection." Anthony discusses a type of interunit connectivity that can help enhance the performance of your neural net application. Expert System cognoscenti can turn to Jay Liebowitz' latest installment of "Managing Expert Systems." Jay reports on national and international academic-world trends in training the next generation of expert system developers.

Aficionados, cognoscenti, literati, gliterati, and fans of American Science will deeply appreciate Technical Editor Hal Berghel's "Cyberspace Navigation." In this article, the esteemed Dr. Berghel cuts a flamboyant path through the Internet -- and displays a rather considerable gift for fiction along the way.

Joseph Schmuller