In suggesting FORTRAN as a computer language for humanists, I am certainly aware that many knowledgeable individuals consider it unsuitable. They point out its general lack of convenient, explicit facilities for handling the alphabetical information that interests humanists, as contrasted with the numerical information that interests humanists, as contrasted with the numerical information that interests scientists. They might observe that the very name stands for FORmula TRANslation, and the mathematical bias implicit in its name and some of its notational features should alone be enough to warn off humanists. While this attitude was certainly valid for the earliest versions of FORTRAN, it is much less true today both of the official standard of the language, ANSI FORTRAN, and the different versions the computer manufacturers make available. After all, FORTRAN was conceived in 1954 to perform the tasks possible on early computers. Today's thirdand fourth-generation computers have capabilities far beyond what was possible then, and FORTRAN as implemented on these machines is a much richer language than it was in 1954.' For example, UNIVAC'S FORTRAN V contains a feature not found in ANSI FORTRAN,jits FLD function,which makes processing alphabetical information very simple. Or again, IBM'S FORTRAN IV level H allows LOGICAL*1 type declarations and certain logical operations which also simplify such processing. In other FORTRANS, ENCODE and DECODE statements serve the same purpose. There is, then a rich diversity among FORTRANs which humanists can exploit for their own purposes. The real problem with FORTRAN is not so much that it will not allow humanists to do what they want but that the way in which they do it will work only on their own or very similar computers. A FORTRAN V program using the FLD function will not work on a CDC or IBM machine while a straightforward numerical program will. In moving such a program from one computer to another, there would undoubtedly be real, though solvable, problems. But those who are skeptical about FORTRAN'S usefulness to humanists ought to keep in mind that Michael Barnett's programming language SNAP, which is so rightly praised for its natural, English-like structure, is actually a disguised version of FORTRAN. The SNAP processor, the program that translates SNAP sentences into instructions the computer can carry out, is written in FORTRAN. Anyone working in SNAP is, as far as the computer is concerned, working in FORTRAN. Nevertheless, though SNAP is excellent as a pedagogical tool, it simply does not have the flexibility and widespread availability of languages like PL/I, FORTRAN, or SNOBOL which make