197204
197204 197204
User Manual: 197204
Open the PDF directly: View PDF .
Page Count: 52
Download | |
Open PDF In Browser | View PDF |
:llENCE & TECHNOLOGY p April, 1972 Vol. 21, No.4- CD COMPUTERIZED OBSERVATORY AT MAUNA KEA, ON THE ISLAND OF HAWAII "Do What I Mean": The Programmer's Assistant Cryptology, The Computer, and Data Privacy Computerizing a Membership Organization The Information Industry and Government Policy The Bad Image That Computers Are Earning Political Lies: An Acceptable Level? W. Teitelman B. Girsdansky R. Pol/ert T. Whitehead W. G. Gearing, and others A. E. Kahn M. W. C. H. _ __IIIIIIIJII!" --- - - - SAN TECH180 1P02121147F6 7301 *N ECHNICAL SERVICES 00808 180 W SAN CARLOS ST SAN JOSE CA 95113 • DO YOU WANT TO PREVENT MISTAKES BEFORE THEY HAPPEN? - avoid pitfalls? - find new paths around old obstacles? - apply in practical situations the observations and wisdom of great scientists and wise men? - stimulate your resourcefulness? - see new solutions to old problems? - distinguish between sense and nonsense? increase your accomplishments? - improve your capacities? IF SOl TRY- The C&A Notebook on COMMON SENSE. ELEMENTARY AND ADVANCED devoted to research, development, exposition, and illustration of one of the most important of all branches of knowledge, i.e. the subject of WHAT IS GENERALLY TRUE AND IMPORTANT = + + + + + + + + THE FIRST SIX ISSUES ARE FREE - see the coupon - THE NEXT 21 ISSUES ARE: Editor: Edmund C. Berkeley, author, businessman, actuary, scientist, computer professional, first secretary of the Association for Computing Machinery 1947-53, editor of Computers and Automation. RETURNABLE IN 7 DAYS FOR FULL REFUND, IF NOT SATISFACTORY WHY NOT TAKE A LOOK? ..... HOW CAN YOU LOSE? - 7. The Elephant and the Grassy Hillside 8. Ground Rules for Arguments 9. False Premises, Valid Reasoning, and True Conclusions 10. The Investigation of Common Sense, Elementary and Advanced 11. Principles of General Science, and Proverbs 12. Common Sense - Questions for Consideration 13. Falling 1800 Feet Down a Mountain 14. The Cult of the Expert 15. Preventing Mistakes from Failure to Understand 16. The Stage of Maturity and Judgment in any Field of Knowledge 17. Doomsday in St. Pierre, Martinique - Common Sense vs. Catastrophe 18. The History of the Doasyoulikes 19. Individuality in Human Beings, ... 20. How to be Silly 21. The Three Earthworms 22. The Cochrans vs. Catastrophe 23. Preventing Mistakes from Forgetting 24. What is Common Sense? - An Operational Definition 25. The Subject of "What is Generally True and Important": Common Sense, Elementary and Advanced 26. Natural History, Patterns, and Common Sense 27. Rationalizing and,Common Sense - - - - - - - - - - - - - - - - - - - - - - - (may be copied on any piece of paper) - - - - - - - - - - - - - - - - - - _ To: COMPUTERS AND A UTOMA TION 815 Washington St., R5, Newtonville, Mass. 02160 YES, please enter my subscription to the C&A Notebook on Common Sense at $12 a year, 24 issues (newsletter style), and extras. Please send me (as FREE premiums for subscribing) the first six issues: 1. Right Answers - A Short Guide to Obtaining Them 4. Strategy in Chess 2. The Empty Column 5. The Barrels and the Elephant 3. The Golden Trumpets of Yap Yap 6. The Argument of the Beard I enclose $ ( ) Please bill me ) Please bill my organization Name ___________________________________ Title_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ Organization ___________________________________________________________ Address ______~------------------------------------------------Signature _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ____ Purchase Order No. ____________ 2 + QUESTIONS AND ANSWERS about "The C&A Notebook on COMMON SENSE, ELEMENTARY AND ADVANCED" INTERESTING: Q: Is the Notebook interesting? NUMBER OF ISSUES.PER YEAR: A: We think so -- but you can judge for yourself. You can see the issues, and if not satisfactory, tell us to discontinue your subscription. Q: I do not want to miss past issues. How do I get them? PAST ISSUES: A: Some of the issues, like "Falling 1800 Feet Down a Mountain" and "Doomsday in St. Pierre, Martinique", are among the most exciting true stories we know. A: Every subscriber's subscription starts at Vol. I, no. 1. Every subscriber eventually receives all issues. Here is how it works. The past issues are sent him four at a time, every week or two, until he has caught up -and thus he does not miss important and interesting issues that never go out of date. Q: Is the Notebook useful? USEFUL: A: It ought to be useful to anybody -- as useful as common sense. There exists no textbook on common sense; the Notebook tries to be a good beginning to common sense, science, and wisdom. UNDERSTANDABLE: Q: How many issues a A: We promise 24 (newsletter style); we anticipate putting out 30 in each yearly volume. Q: Is the Notebook exciting? EXCITING: Q: year do you put out? BOOK: Can I unders tand the Notebook? Q: Are you going to publish all the issues of Volume 1 together as a book? A: No; they do not fit together into a book. A: Yes. It is nontechnical -- written in everyday language and using vivid examples. Q: If I subscribe at the same time to "Computers and Automation", can I receive a premium for subscribing to the Notebook? PREMIUMS: Q: Do you cover in the Notebook all parts of common sense, wisdom, and science in general? COVERAGE: A: Yes, the Four Star Reprint -- see the description below. A: Yes, we plan to. The main subjects so far are: systematic prevention of mistakes; avoiding certain fallacies; important principles; important concepts; illustrative anecdotes; etc. Q: Can I receive the 6 free issues (No. 1 to 6 of the Notebook) without subscribino? A: No. Here is what happens. You subscribe. We send you 8 issues and bill you. You can send them all back in seven days, and the bill is canceled. If you do pay the bill, you receive the 6 free issues, 24 more issues of Vol. 1, and six issues of Vol. 2, no. 1 to 6. Q: Will the Notebook save me from making important mistakes? MISTAKES: A: It ought to. One oft~e main purposes of the Notebook is preventing mistakes. COST: Q: Will the Notebook be worth the cost to me? A: At about 40 cents per issue (30 issues for $12), it is hard for you to lose out. EVEN ONE important mistake prevented, may save you much time, much trouble, and much money. Q: If I do not like the Notebook, can I cancel at any time? GUARANTEE: A: Yes. You will receive a refund for the unmailed portion of ybur subscription. '* * '* * A reprint from CDm~~ FOUR-5TAR REPRINT-------. SCIENCE AND THE ADVANCED SOCIETY, by C. P. Snow, Ministry of Technology, Londoh. England (April, 1966 issue of Computers and Automation). THE INFORMATION REVOLUTION AND THE BILL OF RIGHTS, by Dr. Jerome B. Wiesner, M.I.T. (May, 1971) EMPLOYMENT, EDUCATION, AND THE INDUSTRIAL SYSTEM, by Prof. John Kenneth Galbraith, Harvard Univ. (Aug. 1965) COMPUTERS AND THE CONSUMER, by Rtlph Nader Washington, D.Ci (Oct. 1970) ,- - -(may be copied on any piece of paper)- - TO: Computers and Automation 815 Washington St., Newtonville, Mass. 02160 )YES, you have convinced me to try the Notebook on Common Sense, Elementary and Advanced. Please enter my subscription at $12 a year, 24 issues, newsletter style, and extras. Please send me Issues 1 to 6 as FREE PREMIUMS for subscribing. Please enter my subscription to "Computers and Automation" at the same time ( ) with directory $18.50; ( ) without directory $9.50, and send my FREE premium, the Four Star Reprint. I enclose $._ _ __ Please bill me. Please bill my organization. RETURNABLE IN SEVEN DAYS FOR FULL REFUND IF NOT SATISFACTORY Name,__________________________________________ Title,______________________________________ Organization,________-------------------Address,~--------------------------- Signature_________________P.O. No. ______ COMPUTERS and AUTOMATION for April, 1972 3 Vol. 21, No.4 April, 1972 Editor Edmund C. Berkeley computers and automation Assistant Editors Barbara L. Chaffee Linda Ladd Lovett Neil D. Macdonald Software Editor Stewart B. Nelson Advertising Director Edmund C. Berkeley Art Director RayW. Hass Publisher's Assistant Paul T. Moriarty Contributing Editors John Bennett Moses M. Berlin Andrew D. Booth John W. Carr III Ned Chapin Alston S. Householder Leslie Mezei Ted Schoeters Richard E. Sprague Advisorv Com in ittee James J. Cryan Alston S. Householder Bernard Quint Editon'al0.tftces Berkeley Enterprises,lnc. 815 Washington St., Newtonville, Mass. 02160 617-332-5453 The magazine of the design, applications, and implications of information processing systems - and the pursuit of truth in input, output, and processing. Computers and Programming [T A] 8 "00 WHAT I MEAN": THE PROGRAMMER'S ASSISTANT by Warren Teitelman, Bolt Beranek and Newman, Cambridge, Mass. How to design and implement in a computer, a "programmer's assistant", so that the sometimes mistaken and often experimental behavior of a human programmer is handled with the least waste of his time and effort. 12 CRYPTOLOGY, THE COMPUTER, AND DATA [T A] PRIVACY by M. B. Girsdansky, IBM Research Division, Yorktown, Heights, N.Y. An exploration of how to use the principles and technology of secret or "hidden" writing, in order to secure privacy in computerized records through computer programming. The Computer Industry Advertising Contact THE PUBLISHER Berkeley Enterprises,lnc. 815 Washington St., Newtonville, Mass. 02160 617-332-5453 "Computers and Automation" is published monthly, 12 issues per year, at 815 Washington St., Newtonville, Mass. 02160, by Berkeley Enterprises Inc. Printed in U.S.A. Second Class Postage paid at Boston, Mass. Subscription rates: United States, $9.50 for one year, $18.00 for two years. Canada: add 50 cents a year for postage; foreign, add $3.50 a year for postage. NOTE: The above rates do not include our publication "The Computer Directory and Buyers' Guide"; see "Directory Notice" on the page stated in the Table of Contents. If you elect to receive "The Computer Directory and Buyers' Guide", please add $9.00 per year to your subscription rate. Please address all mail to: Berkeley Enterprises Inc., 815 Washington St., Newtonvi lie, Mass. 02160. Postmaster: Please send all forms 3579 to Berkeley Enterprises Inc., 815 Washington St .. Newtonville, Mass. 02160. © Copyright 1972, by Berkeley Enterprises, Inc. Change of address: If your address changes, please send us both your new address and your old address (as it appears on the magazine address imprint), and allow three weeks for the change to be made. 4 [NT A] MENT POLICY by Clay T. Whitehead, Director, Office of Telecommunications Policy, Executive Office of the President, Washington, D.C. A survey of regulation, competition, and other new forms of economic interaction in the field of communications and computers. [NT A] 21 COMPUTERIZING A MEMBERSHIP ORGANIZATION by William R. Pollert, National Association of Manufacturers, Washington, D.C. How a large membership association planned and implemented a computer facility using a timesharing system. 35 THE MEANING OF AN INTEGRATED DATA [T A] SYSTEM by W. R. Larson, Western Electric, Corporate Education Center, Princeton, N.J. A proposed definition of an integrated data system for an organization, with special attention given to the areas of data storage and retrieval and improvement of system performance. [NT A] 29 THE BAD IMAGE THAT COMPUTERS ARE EARNING by H. W. G. Gearing, Worcester, England - and others Half a dozen examples, in Great Britain and the U.S., of fouled-up computer output and exasperated customers. 24 THE INFORMATION INDUSTRY AND GOVERN- COMPUTERS and AUTOMATION for April, 1972 The Computer Industry (continued) 7 On the Legal Side: THE OUTSIDE DIRECTOR [NT F] by Milton R. Wessel, Attorney, New York, N.Y. 7 "COMPUTERS ENTER THE BUSING CON[NT F) TROVERSY" - ADDENDUM by Robert" L. Glass, Kent, Wash. [NT F] 34 INTERNAL REVENUE SERVICE: USE OF COMPUTERS by William H. Stewart, Jr., University of Alabama 31 IBM'S POWERFUL PARTNER: THE ACCOUNTING [NT F) PRI NCI PLES BOARD Samson Science Corp., New York, N.Y. [NT F) 31 DO YOU WANT TO STOP CRIME? by William P. Wood, III, Richmond, Va. [NT F) 32 CDC VS IBM - CORRECTION by F. O. Parlova, New York, and the Editor 33 Fall Joint Computer Conference: Topics [NT F] Edmund C. Berkeley, Editor, Computen, and Automation 20 "The 1972 Computer Directory and Buyers' Guide" [T GI The Selection of Personnel - for Computers and Other Purposes 26 PICTORIAL REASONING TESTS - PART 5 by Neil Macdonald, Assistant Editor 27 Pictorial Reasoning Test - C&A No.6 [NT F] [NT FI Common Sense, Wisdom, Science in General, and Computers 6 The Old Brain, The New Brain, The Giant Brain and Common Sense by Edmund C. Berkeley, Editor 2 The C&A Notebook on Common Sense 3 Questions and Answers About "The C&A Notebook" [NT E] [NT GI [NT G] Fron t Cover Picture The front cover ~hows the Mauna. Kea Observatory, highest in the world at 13,796 feet above sea level, on the Island of Hawaii on the top of an The 88-inch teleinactive volcano. scope and the dome aperture are controlled by an IBM 1800 data acquisition and control system. The system now responds automatically to star motion and the air idistortion along the line of sight, and will eventually respond automatically to temperature, humidity, and wind speed. The Profession of Information Engineer and the Pursuit of Truth 37 DALLAS: WHO, HOW, WHY? - Part II [NT A] by Mikhail Sagatelyan, Moscow, USSR A report published in Leningrad, USSR, by an ace Soviet reporter about the circumstances of the assassination of President John F. Kennedy and their significance from a Soviet point of view: Part 2. 44 Political Lies: An Acceptable Level? [NT Al by Richard M. Nixon, J. W. Fulbright, and others 33 A Concerted Campaign To Deny the American People [NT G] Essential Knowledge About the Operation of Their Government by Prof. Henry Steele Commager, Amherst, Mass. 32 Reducing and Dismantling Science and Research [NT G) Institutions, and Social Responsibility by A. G. Michalitsanos, Cambridge, England 7 Dealing With Today's Problems [NT F) by John Skowronski, OPMA, Kansas City, Mo. 34 Some Responsibility for Our Chaotic Society [NT G] by S. R. Harrison, Trondheim, Norway 34 Unsettling, Disturbing, Critical ... [NT FI Statement of policy by "Computers and Automation" Departments 49 20 Advertising Index Calendar of Coming Events Correction Monthly Computer Census New Contracts New Installations 32 48 46 47 Key [A] [C] [E] [F] [G] [NTI [T] Article Monthly Column Editorial Forum The Golden Trumpet Not Technical Technical Computers, Games, and Puzzles 36, 43 Numbles, by Neil Macdonald 49 Problem Corner, by Walter Penney, COP COMPUTERS and AUTOMATION for April, 1972 [T C) [T C] 5 C- a EDITORIAL The Old Brain, The New Brain, The Giant Brain, and Common Sense All computer people from time to time think about brains - those interesting and complex devices made out of common chemicals that we human beings and other animals carry around inside our heads and by means of which we think. The prototype brain began in the course of evolution more than 400 million years ago. As mammals and birds developed, their brains became very efficient instruments. These brains accept thousands of simultaneous impressions from one or more senses, and make use of a builtin yet trainable evaluating function. The evaluating function uses observations, context, surroundings, instincts, learned experiences, and other information in order to make decisions. The decisions are the ordinary everyday decisions that are useful and relevant to solving the ageold problems of surviving: • How to recognize and avoid danger; • How to find and choose food and shelter; • How to satisfy other wants, how to mate, etc. This b rain is what I like to call "the old brain". It is present, of course, in all human beings. Two of its most remarkable parts are: the way it deals with streams of multiple input sensations; and the evaluating function. Together these parts enable the animal to decide swiftly and almost automatically what to do next and how fast he or she should do it. The output of the old brain is a great quantity of "common sense" behavior - consisting mainly of directions to the body for survival purposes. At some point more than one million years ago (and probably not more than 10 million years ago) a branch of the primates began to develop a new brain. The new brain has produced language, words to refer to ideas, tool making, culture, the invention of symbols and writing, the solving of problems by using symbols, and much more. The new brain is able to analyze problems into many logical steps, to talk about these steps, to imagine these steps in different arrangements. Most extraordinary of all, the new brain possesses the equipment (unlike any other animal except perhaps the dolphin) to deal with the meanings of tens of thousands of words. The output of the new brain, when brought up in an environment of education and civilization, is intellectual behavior that no other species of animal can come close to producing. The origin of the new brain is illuminated in the following quotation from "The Naked Ape" by Desmond Morris (a fascinating book): Of all the non-specialists, the monkeys and the apes are perhaps the most opportunist. As a group, they have specialized in non-specialization. And among the monkeys and apes, the naked ape is the most supreme opportunist of them all. ... All young monkeys are inquisitive but the intensity of their curiosity tends to fade 6 as they become adult. With us the infantile inquisitiveness is strengthened and stretched into our mature years. We never stop investigating. We are never satisfied that we know enough to get by. Every question we answer leads on to another question. This has become the greatest survival trick of our species. For the opportunist the going may always be rough, but the creature will be able to adapt quickly to any quick-change act that the environment decides to put Ion. The new brain, in spite of its power, has some severe limitations. It seems to be able to pay attention to only one idea at a time, while the old brain seems to be able to pay attention to many aspects of the environment simultaneously and tJ react almost immediately to any suddenly disturbing event, like something noticed out of the comer of one's eye. The new brain makes far more mistakes than the old brain. The new brain seems to have rather a small memory - at least, only when human beings make use of external recording can they remember' more than a small part of what they consciously want to remember. The old brain in contrast seems to have a very large amount of information ready for use, especially in emergencies. And the new brain seems not to have access to all the information stored in the old brain. The giant brain of course is the computer. The phrase has been applied to computers since 1949. The giant brain is strictly an imitation of the new brain. It is not at all an imitation of the old brain - because it does not automatically receive thousands of sense impressions per second; it does not automatically integrate these impressions into representations of the environment; it does not automatically operate an evaluating function which solves the problem of living and surviving from moment to moment: Like the new brain, the giant brain deals with symbols; it performs operations of logic and arithmetic in enormous quantities, and at a speed that is on the order of a million times the speed of human beings. "Common sense" behavior is what an animal displays when he is using his old brain: the sense to avoid danger, to find and choose food, to survive according to the bag of survival tricks developed by his species. Computers are enormously busy in these days performing the functions of the new brain. We are just beginning to apply computers to performing the functions of the old brain - as in optical character recognition. It is time to pay more attention to the construction of computers that will perform adequately the functions of the old brain, and produce "common sense" behavior. Editor COMPUTERS and AUTOMATION for April, 1972 On The Lega I Side: THE OUTSIDE DIRECTOR Milton R. Wessel, Attorney New York, N. Y. A person who accepts election as an "outside director" of a public company merely for honor or prestige, is making a serious mistake. During the last decade a host of litigations have made very clear that a director -- outside or inside -- has affirmative personal responsibility to a large number of persons, including stockholders, employees,. suppliers, customers, and the public generally. Often the standard corporate indemnification arrangement will not help him where he fails to perform that responsibility, and he must pay money damages and in extreme cases even suffer criminal conviction. The fact that the organization involved is a church, hospital, college or other non-profit organization does not relieve him of even a shred of his responsibility. What is the director's responsibility? A complete answer requires a book, and much has been written. The short, practical answer here is that a director must act reasonably under the circumstances; he must act as one would expect a director to act. Section 717 of the New York Business Corporation Law puts it thus: "Directors and officers shall discharge the duties of their respective positions in good faith and with that degree of diligence, care, and skill which ordinarily prudent men would exercise under similar circumstances in like posi tions." All of this means that the outside director must pay sufficient attention to the business to satisfy his responsibility. He cannot successfully defend against liability by saying, "I didn't know" or "I didn't attend meetings" or "They wouldn't give me the financial data I asked for". For example, a director of a computer company, whose programs and banks of data are its main asset, should satisfy himself that they are reasonably protected against fire and fraud. If he does not and the company is wiped out, he may be liable to shareholders for negligence. He can't just sit and wait for management to ask for his approval of a security program. Nor can a director avoid liability by closing his eyes to facts about which he should know. If he hears something at a board meeting which he doesn't like, he can't say, "Don't discuss that here" or "Don't tell me about it." He must be able honestly to say that in his judgment, the company is doing the best it can under all circumstances. If he cannot, he must let the other directors know what is wrong in his opinion. If they do not agree and cannot persuade him to the contrary, he must withdraw promptly; he cannot sit and wait until he spots trouble just ahead. This column deals only with the affirmative duties of a public director to be a reasonably effective one. In addition, there are a host of other somewhat more specific prohibitions which are of an essentially negative character and which are far more precisely defined (not to disclose confidences; not to profit by corporate opportunities; not to engage in short swing or other improper insider trading, -- in short, not to deal unfairly with the company), which need discussion. The result of these two sets of increased obligations is that informed businessmen and professionals COMPUTERS and AUTOMATION for April, 1972 are more and more reluctant to accept appointment to public boards without compensation commensurate to the degree of responsiblity assumed, or a major stake in the enterprise, or (more usually) both. I predict this trend will continue until all directors finally recognize their responsibilities and boards become true and effective management units, rather than the dummy operations which characterize so many newly public companies. "COMPUTERS ENTER THE BUSING CONTROVERSY" - ADDENDUM Robert L. Glass 26414124th Ave., SE Kent, Wash. 98031 For the sake of journalistic completeness, I am sending this note on Seattle's program of busing for integration as discussed in my article "Computers Enter the Busing Controversy" which was published in your July issue. My statement "Poli tics and public outcry may yet negate the integration plan which technology has been able to produce" was, unfortunately, prophetic. A last-minute local court decision blocked busing and nearly scuttled Seattle's entire middle school concept. Schools opened, but busing itself was postponed (nominally) a year. Citizens Against Mand~tory Busing (CAMS), instigator of the suit which led to the decision, was jubilant. The School Board, with too little time left for an effective countermove, could only acknowledge temporary defeat. Busing advocates in Seattle, and the computing technologists who constructed the busing implementation plan, can only fall back on the traditional loser's cry, "Wait 'til next year." DEALING WITH TODA Y'S PROBLEMS John Skowronski Director of Publications Kansas City Chapter, DPMA P. O. Box 2425 Kansas City, Mo. 64142 I would like your permission to reprint the article "The Handwriting on the Wall" (June, 1971, p. 6) which editorializes a piece entitled "Computerized" by "Margo". I am also writing to the Boston Globe for their permission to reprint the "Margo" article. Data Processing Management Association (D.P.M.A.) as you probably know is a non-profit organization of data processing professionals with approximately 260 members in the Kansas City Chapter. Our local newsletter has a monthly distribution of approximately 300. I would also like to congratulate you for a very fine magazine that deals with today's problems. Your articles "The Case of Secret Service Agent Abraham W. Bolden" (June, 1971, p. 41) and "The Issue is Hypocrisy" (June~ 1971, p. 45) were quite revealing. My wife was so impressed that she is going to subscribe personally for your magazine. 7 1100 WHAT I MEAN II : Th e p tA rogrammer 5 ssistant Warren Teitelman Bolt Beranek and Newman 50 Moulton St. Cambridge, Mass. 02140 "The programmer's environment influences, and to a large extent determines, what sort of problem he can tackle, and how far he can go in a given time." This article deals with the design and actual implementation in a computer programming system of "a programmer's assistant". The general function of the "programmer's assistant" is to make it possible for the human programmer to say to the computer "do what I mean" instead of "do what I say", and "undo what I just tried - it did not work", instead of leaving the programmer with the sad consequences of his actual instructions. In other words, the programmer's assistant deals with such factors as: • ease of interaction; • level of interaction; • forgiveness for errors (both spelling errors and errors of thought); • going back and taking a different path; • changing one's mind; etc. and in general, the programmer's environment. This area of improvement in interactive programming is important. For many applications, the programmer's environment influences, and to a large extent determines, what sort of problem he can tackle, and how far he can go in a given time. If the "environment" is "cooperative" and "helpful", then the programmer can be more ambitious and productive. If not, he may spend much of his time and energy performing routine clerical tasks and "fighting the system". An Analogy The conceptual role of the programmer's assistant can perhaps best be explained by using an analogy with a well-equipped chemistry laboratory. We are attempting to recreate for the programmer the type of laboratory assistant who works closely with the chemical scientists. As an example of his duties, the assistant can watch the scientist perform a sequence of operations, and then be instructed to repeat the operation, perhaps with some rearrangements, omissions, or other modifications, or to clean up the mess and make things ready for another run. This kind of general purpose assistance complements the facilities in the laboratory, but does not of course substitute for them. Its value is greatest in those situations which are (1) not standardized, so that special equipment has not already been built for that purpose, or (2) not sufficiently repetitive, so that new equipment cannot be designed and built cost-effectively. These situations usually require the scientist to achieve 8 his goal by executing many small steps, some similarities or repetitions, while each point what to do next on the basis vious results, with occasional, or even "backing and filling." with perhaps deciding at of the prefrequent, In the realm of programming, it is easier to build and discard tools for just a few applications than in the physical sciences. However, all too frequently the programmer may not realize he is going to need an operation repeated, or else the ·operation is not exactly the same in each case, or else the effort of building the tool (and debugging it) may not be worth interrupting his train of thought and action. If it were convenient for him to specify repeating a previous operation, including possible modifications, without his having previously prepared for it, he would certainly do so. Then like our hypothetical scientist, he would be able to turn his attention to evaluating the resuI ts of his "experiment" so fa r, and deciding what to do next, while his assistant performs the dog work. Undo One of the important functions of the laboratory assistant is cleaning up an abortive experiment and getting things ready for the next run. In the realm of programming, users prepare for possible disasters such as a program running wild and chewing up a data structure, by saving the state of part or all of their environment before attempting some not yet reliable process. Undoing then consists of backing up to some previous state and starting over from there. However, this saving and dumping operation is usually expensive and time-consuming. Of course disaster may also strike as a result of supposedly innocuous operation. As a result, the user all too often finds himself in a situation where he wishes, either idly or desperately, that he could reverse the effect of a previous operation or operations. Here we have an advantage over the physical sciences: a laboratory assistant may not be able to reverse the effect of mixing two chemicals together, no matter how hard or skillfully he tries. However, in the programmer's assistant, we can provide such a capability. BBN-LiSP At Bolt, Beranek, and Newman, Cambridge, Mass., and at over a dozen other installations around the United States, the language called BBN-LISP is in COMPUTERS and AUTOMATION for April, 1972 common use. BBN-LISP as a programming language, is an implementation of LISP, a language designed for list processing and symbolic manipulation. BBN-LISP as a programming system, is the product of, and vehicle for, a research effort supported by the Advanced Research Projects Agency for improving the programmer's environment. Design Philosophy BBN-LISP contains many user-support facilities. These include: • a sophisticated structure editor which can either be used interactively or as a subroutine; • a debugging package for inserting conditional programmed interrupts around or inside of specified procedures; • a "prettyprint" facility for producing structured symbolic output; • a program analysis package which produces a tree-structured representation of the flow of control between procedures; • a concordance listing indicating for each procedure the procedures that call it, the procedures that it calls, and the variables it references, sets, and binds; etc. Most on-line programming systems contain similar features. But the essential difference between BBN-LISP and other systems is embodied in the philosophy that the user addresses the system through an (active) intermediary agent, whose task it is to collect and save information about what the user and his programs are doing, and to utilize this information to asist the user and his programs. This intermediary has been named the programmer's assistant (or p.a.). The Programmer's Assistant For most interactions with the BBN LISP system, the programmer's assistant is an invisible interface between the user and LISP: the user types a request, for example, specifying a function to be applied to a set of arguments; the indicated operation is then performed, and a resulting value is printed. The system is then ready for the next request. However, in addition, in BBN-LISP, each input typed by the user, and the value of the corresponding operation, are automatically stored by the p.a. on a global data structure called the history list. The history list contains information associated with each of the individual "events" that have occurred in the system, where an event corresponds to an individual type-in operation. Associated with each event is the input that initiated it, the value it yielded, plus other optional information such as side effects, messages printed by the system or user programs, information about any errors that may have occurred during the execution of the event, etc. As new events occur, existing events are aged, and the oldest event or events are "forgotten." The length of the history list is set by the user, and is typically somewhere between 30 and 100 events. The user can refer to an event on the history list by its relative event number, for example, -1 refers to ~he most rece~t event, -2 the event before that, etc. He may refer by an absolute event number, or by a pattern which is then used for searching the history list. For example, the user can retrieve an event in order to REDO a test case after making some program changes. Or, having COMPUTERS and AUTOMATION for April, 1972 typed a request that contains a slight error, the user may elect to FIX it, rather than retyping the request in its entirety. The p.a. recognizes such requests as REDO and FIX as being directed to it, not the LISP interpreter, and executes them directly. For example,when given a REDO command, the p.a. retrieves the indicated event, obtains the input from that event, and treats it exactly as though the user had typed it in directly. Similarly, the FIX command directs the p.a. to allow the user to edit the indicated input. When the user has fixed up this input, it is again processed by the p.a. exactly as though it had been typed in. Fifteen Commands The p.a. currently recognizes about 15 different commands (and includes a facility for the user to define additional ones). For example, the USE command provides a convenient way of specifying simultaneous substitutions for lexical units and/or character strings, e.g. USE X FOR Y AND + FOR *. This permits after-the-fact parameterization of previous events. The p.a. also enables the user to treat several events as a single unit, (e.g. REDO 47 THRU 51), and to name an event or group of events. All of these capabilities allow, and in fact encourage, the user to construct complex console operations out of simpler ones in much the same fashion as programs are constructed, i.e. simpler operations are checked out first, and then combined and rearranged into large ones. The important point to note is that the user does not have to prepare in advance for possible future use or reuse of an event. He can operate straightforwardly as in other systems; yet the information saved by the p.a. enables him to implement his "after-thoughts." Undoing Perhaps the most important after-thought operation made possible by the p.a. is that of undoing the side-effects of a particular event or events. In most systems, if the user suspects that a disaster might result from a particular operation, e.g. an untested program running wild and chewing up a complex data structure, he would prepare for this contingency by saving the state of part or all of his environment before attempting the operation. If anything went wrong, he would then back up and start over. Such "accidents" happen all too often in typical console sessions, and result in the user's either having to spend what may involve a great deal of effort in rectinstructing the inadvertently destroyed information, or alternatively in returning to his last back-up, in which case the user would have to redo any useful work performed since that backup. Instead with the p.a., the user can recover by simply typing UNDO, and then perform the originally intended operation. The existence of UNDO frees the user frqm worrying about such oversights. He can be relaxed and confident in his console operations, yet still work rapidly. He can even experiment with various program and data configurations, without necessarily thinking through all the implications in advance. One might argue that this would promote sloppy working habits. However, the same argument can be and has been leveled against interactive systems in general. In fact, fteeing the user from having to anticipate all of th~ consequences of an (experimental) change usually results in his being abl~ to 9 pay more attention to the conceptual difficulties of the problem he is trying to solve. Another advantage of undoing as it is implemented in the programmer's assistant is that it enables events to be undone selectively. Finally, since the operation of undoing an event itself produces side effects, it too is undoable. The user can often take advantage of this fact, and employ strategies that use UNDO for desired operation reversals, not simply as a means of recovery in case of trouble. For example, suppose the user wishes to interrogate a complex data structure in each of two states while successively modifying his programs. He can interrogate the data structure, change it, interrogate it again, undo the changes, modify his programs, and then repeat the process using successive UNDOs to flip back and forth between the two states of the data structure. 1m plementati on The UNDO capability of the programmer's assistant is implemented by making each function that is to be undoable save on the history list enough information to enable reversal of its side effects, For example, when a list node is about to be changed, it and its original contents are saved. For each primitive operation that involves side effects, there are two separate functions, one which always saves this information, i.e., is always undoable, and one which does not save it. errors, especially those occurring in type-in, are of the type that can be corrected without any knowledge about the purpose of the program or operation in question, e.g. misspellings, certain kinds of syntax errors, etc. The p.a. attempts to correct these errors, using as a guide both the program's environment at the time of the error, and information gathered by monitoring the user's requests. This form of implicit assistance provided by the programmer's assistant is named the DWIM (Qo-~hat-l-Mean) capability. DWIM is also used to correct other types of conditions not considered errors, but nevertheless obviously not what the user meant. For example, if the editor is called on a function that is not defined, rather than typing NOT EDITABLE, the editor invokes the spelling corrector to try to find what function the user meant to edit, giving DWIM as possible candidates a list of user-defined functions. Similarly, the spelling corrector is called to correct misspelled edit commands, p.a. commands, names of files, etc. The spelling corrector can also be called by user programs. As mentioned above, DWIM also uses information gathered by monitoring user requests. This is accomplished by having the p.a., for each user request, "notice" the functions and variables being used, and add them to appropriate spelling lists, which are then used for comparison with (possibly) misspelled units. Thus, DWIM "may know" that FACT 'was the name of a function, and is therefore able to correct FATC to FACT. Although the overhead for saving undo information is small, the user may elect to make a particular operation not be undoable if the cumulative effect of saving the undo information seriously degrades the overall performance of a program, because the operation in question is repeated so often. The user, by his choice of function specifies which operations are undoable. In a sense, the user's choice of function acts as a declaration about frequency of use versus need for undoing. For those cases where the user does not want certain functions undoable once his program becomes operational, but does wish to protect himself against malfunctioning while debugging, the p.a. provides a facility called TESTMODE. When in TESTMODE, the undoable version of each function is executed, regardless of whether the user's program specifically called that version or not. As a result of knowing the names of user functions and variables (as well as the names of the most frequently used system functions and variables), DWIM seldom fails to correct an error the user feels it should have. And since the spelling corrector knows about common typing errors, e.g. transpositions, doubled characters, shift and case mistakes, etc., DWIM almost never mistaken!y corrects an error. However, if DWIM did make a mistake, the user could simply interrupt or abort the computation, UNDO the correction (all DWIM corrections are undoable), and repair the problem himself. Since an error had occurred, the user would have had to intervene anyway, so that DWIM's mistaken correction did not result in extra work for him. It is this benign quality that makes DWIM so appreciated by users. Side Effects Example Finally, all operations involving side effects that are typed-in by the user are automatically made undoable by the p.a. by substituting the corresponding undoable function name in the expression before evaluation, as in the earlier example with the REMOVE PROPERTY operation. This procedure is feasible because operations that are typed-in rarely involve iterations or lengthly computations directly nor is efficiency usually important. However, as a precaution, if an event occurs during which more than a user-specified number of pieces of undo information are saved, the p.a. interrupts the operation to ask the user if he wants to continue having undo information saved. Automatic Error Correction - The DWIM Facility The previous discussion has described ways in which the programmer's assistant is explicitly invoked by the user. The programmer's assistant is also called when certain error conditions are encountered. A surprisingly large percentage of these 10 Suppose t::e user defines a function FACT of one argument, N. The value of FACT[N] is to be N factorial. (LAMBDA (N) (COND ((ZEROP N9 1) ((T (ITIMS N (FACCT 8SUBl N] (FACT) ~DEFINEQ((FACT Note that the definition of FACT contains several mistakes: ITIMES and FACT have been misspelled. The 9 in N9 was intended to be a right parenthesis, but the teletype shift key was not depressed; similarly, the 8 in 8SUBl was intended to be a left parenthesis; and finally, there is an extra left parenthesis in front of the first T, according to LISP, on the second line. ~PRETTYPRNT((FACCT] [1J =PRETTYPRINT =FACT [3J [2J COMPUTERS and AUTOMATION for April, 1972 (FACT [LAMBDA (N) (COND « ZEROP N9 1) «T (ITIMS N (FACCT 8SUBI NJ) NIL ~FACT(3) Now, after defining FACT, the user wished to look at its definition using PRETTYPRINT, which he unfortunately misspells, [IJ. Since there is no function PRETTYPRNT in the system, an "undefined function" error occurs, and DWIM is called. DWIM invokes its spelling corrector, which searches a list of functions frequently used (by this user) for the best possible match. Finding one that is extremely close, DWIM proceeds on the assumption that PRETTYPRNT meant PRETTYPRINT, notifies the user of this, [2J and calls PRETTYPRINT. At this point, PRETTYPRINT would normally print (FACCT NOT PRINTABLE) and exit, since FACCT has no definition. Note that this is not a LISP error condition, so that DWIM would not ~called as described above. However, it is obviously not what the user meant. This sort of mistake is corrected by having PRETTYPRINT itself explicitly invoke the spelling corrector portion of DWIM whenever given a function with no satisfactory definition. Thus with the aid of DWIM, PRETTYPRINT is able to determine that the user wants to see the definition of the function FACT, [3J and proceeds accordingly. ~FACT(3) [ 4J N9(IN FACT) »--> N) (IN FACT) (COND -- «T --))) »--> (COND -- (T --)) ITIMS(IN FACT)->ITIMES [5J FACCT(IN FACT)->FACT 8SUBl(IN FACT) »--> (SUBI U.B.A. N9(IN FACT) FIX? YES N9(IN FACT) »--> N) U.D.F. TON FACT) FIX? YES (COND -- «T --))) »--> (COND (T --)) ITIMSON FACT) ->ITIMES? ... YES FACCT(IN FACT) -:>FACT? ... YES U.B.A. 8SUB1(IN FACT) FIX? NO U.B.A. (8SUBI BROKEN) [lJ [2J [3] [4J [5J A great deal of effort has been put into making DWIM 'smart'. Experience with perhaps a dozen different users indicates we have been very successful; DWIM seldom fails to ~orrect an error the user feels it should have, and almost never mistakenly corrects an error. However, it is important to note that even when DWIM is wrong, almost always no harm is done; since an error had occurred, the user would have had to intervene anyway if DWIM took no action. Thus, if DWIM mistakenly corrects an error, the user simply interrupts or aborts the computation, UNDOes the DWIM change using UNDO, and makes the correction he would have had to make without DWIM. It is this benign quality of DWIM that makes it a valuable part of the programmer's assistant. Conclusion When a user types a request which contains a misspelling, having to retype it is a minor annoyance (depending, of course, on amount of typing required and the ~ser's typing skill). However, if the user has mentally already performed that task, and is thinking ahead several steps to what he wants to do next, then having to go back and retype the operation represents a disruption of his thought processes, in addition to being a clerical annoyance. The disruption is even more severe when the user must also repair the damage caused by a faulty operation, instead of simply UNDOing. 6 ~PP FACT [ 6J (FACT [LAMBDA (N) (COND «ZEROP N) 1) (T (ITIMES N (FACT (SUBI NJ) NIL The user now calls his function FACT, [4J. During its execution, five errors occur, and DWIM is called fi ve times, [5J. At each poi nt, the error is corrected, a message printed describing the action taken, and the computation allowed to continue as if no error had occurred. Following the last correction, the value of FACT (3) is printed. Finally, the user prettyprints the new, now correct, definition of FACT, [6J. In this particular example, the user was shown operating in TRUSTING mode, which gives DWIM carte blanche for all corrections. The user can also operate in CAUTIOUS mode, in which case DWIM will inform him of intended corrections before they are made, and allow the user to approve or disapprove of them. For most corrections, if the user does not respond in a specified interval of time, DWIM automatically proceeds with the correction, so that the user need intervene only when he does not approve. Sample output is given below. Note that the user responded to the first, second, and fifth questions; DWIM responded for him on the third and fourth. COMPUTERS and AUTOMATION for April, 1972 We have found that in addition to making life a lot more pleasant for users, the p.a. has a synergistic effect on user productivity. This effect seems to be related to the overhead that is involved when people have to switch tasks or levels: The DWIM facility greatly facilitates construction of complex programs because it allows the user to remain thinking about his program operation at a relatively high level without having to descend into manipulation of details. Even more important, however, is that DWIM and thep.a. act to minimize user distractions and diversions at whatever level the user is operating or chooses to operate. We believe that p.a. facilities should be built into low-level debugging packages, the executive language of time sharing systems, etc., as well as other 'high-level' programming languages, for these facilities provide the user with a significant 'mental mechanical advantage' in attacking problems. 0 References 1. Teitelman, W., Bobrow, D.G., Hartley, A.K., and Murphy, D.L., "BBN-LISP TENEX Reference Manual", BBN Report, July 1971. 2. Tei telman, W., "Toward a Programming Laboratory", IJCAI, January 1969. 3. Bobrow, D.G., "The Future of List Processing Languages", to be published in a forthcoming issue of the "Communications of the ACM". 11 Cryptology, The Computer, and Data Privacy M. B. Girsdansky Scientific Information Dept. I BM Research Division P.O. Box 218 Yorktown Heights, N. Y. 10598 "In the modern digital computer, cryptology has gained a valuable ally, but in ever-growing measure, the computer may find itself in need of cryptology. " English writer and wit G. K. Chesterton once observed that a nighttime view of neon-lit Broadway would be singularly beautiful-to someone unable to read. Implicit in the quip was 'realization that any alphabetic (or syllabic) system of writing in. fact constitutes a cipher to one ignorant of the system. * An unfamiliar script long concealed the language and content of the ancient writings known as Minoan Linear B. It was only in the early 1950s that cipher-breaking techniques identified the material as Greek of the Late Bron,ze Age (c. 1300-1000 B. C.), radically revising our notions of early Mediterranean history. Such "hidden writing" (the etymological meaning of "cryptography") is the result of historical accident, but there are examples of intentional cryptography which are even older. Deliberately enciphered hieroglyphics were inscribed more than half a millennium before Linear B was set down. Closer to our own day-in the India of the Fourth Century B. C.-ambassadors to foreign courts were explicitly advised to practice "the decipherment of secret writings." Over the centuries since, cryptology-the discipline concerned with the making and breaking of codes and ciphers-has led a sometimes raffish, but usually important and frequently fascinating existence. Like much else, cryptology has been profoundly affected by technology. Telegraph, telephone, radio, and a variety of subsequently developed means of processing information electronically have presented problems as well as opportunities. Thus, encoded or enciphered messages had to be made far more resistant to garbling * A cipher transforms low·order constituents of a message-such as individual characters or, more rarely, small groups of characters-according to some definite strategy. In contrast, a code consists of an agreed-upon but essentially arhitrary listing of natural-language segments vs. code equivalents, allowing a small group of symbols to represent an entire word, sentence, or even group of sentences. A volume containing such entries as DA B L I = "Consign ment two weeks overdue/Customer threatens cancellation/Expedite or give particulars" is a true code book. The socalled Morse Code is a misnomer; actually tht:- system is a form of simple substitution cipher. (Based on a report published in 'IBM Research Reports', Vol. 7, No.4, 1971, and reprinted with permission) 12 over noisy lines or during poor atmospheric conditions. New techniques for encrypting data (e. g., such novelities as high-speed transmission and signal-scrambling) were devised and in turn had to be countered. Increasingly, cryptoJr ' drew upon the insights and skills of mathematicians, electrical engineers, and others from the physical sciences. Of perhaps greatest ultimate significance at the cryptology/technology interface was development of the modem digital computer. In it, cryptology has gained a valuable ally, but in ever-growing measure, the computer may well find itself in need of cryptology. As the computer is used to store and process data on individuals, concern has been voiced lest much that is personal and legitimately private lie open to those with access to "data banks." In response, cryptologic techniques have been suggested as a logical means of shielding a computer's contents, which might otherwise be readily called out from the machine. Depending on the purposes to which the data bank is to be put, some or all of its contents should be inaccessible and unreadable to anyone but a given authorized user. Granted, just who is to be "an authorized user" is largely a question of public policy. Exploration of technical possibilities is, however, the province of the scientist and the engineer. In this area, workers at the IBM Research Division have been among those seeking effective privacy systems, as well as probing the weaknesses of systems already proposed. If, as seems likely, data banks are to be embodied as centralized computers reached through a network of remote terminals, it would be desirable to have many features of the system-not only a cipher itself-work together to provide privacy and security from tampering. Such an approach has been evolved by IBM's Horst Feistel, and implemented in an experimental system developed by William A. Notz and J. Lynn Smith which features critical hardwar~ designed by Smith. Perhaps logically prior to system details, however, is the matter of what type of cipher to use. It has long been realized that certain classes of ciphers are, by their very nature, weaker than others. At IBM, Bryant Tuckerman has gone somewhat further, showing that one popular COMPUTERS and AUTOMATION for April, 1972 Plaintext COMEHERE Key Ciphertext EXITEXIT GLUXLBZX tttttttt ... ... ... ... ... ... ABC D E F CHI J K L M N 0 P A B C D -.E F G H -.J J K L M N o p Q R S -+T U V II -+X y a I 2 3 4 5 6 7 8 9 10" 12 13 14 15 16 17 18 19 3J 21 22 23 24 25 ABC D E F CHI J K L M N 0 P Q R STU V II X Y Z ... S Q R T U V II X Y Z ABC D E F CHI J K L M N 0 P Q R STU V II X Y Z BCD E F CHI J K L M N 0 P Q R STU V II X Y Z A C D E F CHI J K L M N 0 P Q R STU V II X Y Z A B D E F CHI J K L M N 0 P Q R STU V II X Y Z ABC E F CHI J K L M N 0 P Q R STU V II X Y Z ABC D F CHI J K L M N 0 P Q R STU V W x Y Z ABC D E CHI J K L M N 0 P Q R STU V W x Y Z ABC D E F H I J K L M N 0 P Q R STU V W x Y Z ABC D E F G J J K L M N 0 P Q R STU V W x Y Z ABC D E F G H J K L M N 0 P Q R STU V W x Y Z ABC D E F C H J K L M N 0 P Q R STU V W x Y Z ABC D E F CHI J L M N 0 P Q R STU V W x Y Z ABC D E F CHI J K M N 0 P Q R STU V W x Y Z ABC D E F C H J J K L N 0 P Q R STU V W x Y Z ABC D E F CHI J K L M o P Q R STU V W x Y Z ABC D E F G H I J K L M N P Q.R STU V W x Y Z ABC D E F G H I J K L M N 0 Q R STU V W x Y Z ABC D E F G H I J K L M N 0 P R STU V W x Y Z ABC D E F G H I J K·L M N 0 P Q STU V W x Y Z ABC D E F CHI J K L M N 0 P Q R T U V II X Y Z ABC D E F CHI J K L M N 0 P Q R S U V II X Y Z ABC D E F G H J J K L M N 0 P Q R S T V II X Y Z ABC D E F CHI J K L M N 0 P Q R STU II X Y Z ABC D E F G H I J K L M N 0 P Q R STU V X Y Z ABC D E F G H I J K L M N 0 P Q R STU V W Y Z ABC D E F CHI J K L M N 0 P Q R STU V W X Z ABC D E F G H I J K L M N 0 P Q R STU V W X Y oA I B 2 C 3 D 4 E 5 F 6 G -+7 H 8 J ABCDEFGHJJKLMNOPQRSTUVWXYZ BCD E F G H I J K L M N 0 P Q R STU V II X Y Z A C D E F G H I J K L M N 0 P Q R STU V W X Y Z A B D E F G H I J K L M N 0 P Q R STU V W X Y Z ABC E F G H I J K L M N 0 P Q R STU V II X Y Z ABC D F CHI J K L M N 0 P Q R STU V W X Y Z ABC D E G H I J K L M N 0 P Q R STU V II X Y Z ABC D E F H I J K L M N 0 P Q R STU V W X Y Z ABC D E F C I J X L M N 0 P Q R STU V II X Y Z ABC D E F C H 9 J J K L M N 0 P Q R STU V W X Y Z ABC D E F CHI 10 K X L M N 0 P Q R STU V W X Y Z ABC D E F G H I J "L L M N 0 P Q R STU V II X Y Z ABC D E F G H I J K ~ M M N 0 P Q R STU V W X Y Z ABC D E F G H I J K L g N N 0 P Q R STU V W X Y Z ABC D E F CHI J K L M N 0 0 P Q R STU V II X Y Z ABC D E F G H I J K L M N ~ P P Q R STU V W X Y Z ABC D E F G H I J K L M N 0 ~ Q Q R STU V W X Y Z ABC D E F G H I J K L M N 0 P "R R STU V II X Y Z ABC D E F CHI J X L M N 0 P Q ffi S S T U V W X Y Z ABC D E F G H I J K L M N 0 P Q R ffi T T U V II X Y Z ABC D E F CHI J X L M N 0 P Q R S m U U V W X Y Z ABC D E F C H ~ J K L M N 0 P Q R S T 21 V V II X Y Z ABC D E F G H I J K L M N 0 P Q R STU 22 W II X Y Z ABC D E F G H I J K L M N 0 P Q R STU V 23 X X Y Z ABC D E F G H I J K L M N 0 P Q R STU V W ~ Y Y Z ABC D E F G H I J K L M N 0 P Q R STU V II X 3 Z Z ABC D E F G H I J K L M N 0 P Q R STU V W X Y Figure 1. "Classic Vigenere" enciphered by establishing a oneto-one correspondence between plaintext and key characters. The key is repeated in whole or in part when necessary. Ciphertext characters are then those which stand at the intersections of key rows and plaintext columns. Figure 2. Vigenere encipherment considered as addition modulo 26. If the letters of the alphabet are given the values A = 0, ... Z = 25, and combination of a row and a column is defined as "addition," then the character at the intersection of row and column is the sum, modulo 26. Here, 21 + 7 = 2, corresponding to V + H = C. class of ciphers is, indeed, far weaker than even a skilled amateur at cryptology might think. of the plaintext and of the key. For mnemonic ease, it was formerly customary for the key to consist of an intelligible word or meaningful phrase, although this feature constitutes a weakness; incoherent keys, which reduce the number of clues, are certainly preferable. VIGENERE-VERNAM CIPHERS The cipher class studied by Tuckerman occupies a classic position in traditional cryptology. Although some members of the class are centuries old, one variety was devised a matter of decades ago to provide on-line encipherment of teletypewriter text. The class has been proposed by some as a means of furnishing privacy in modern data storage and transmission systems. Both theoretically and by practical demonstration, Tuckerman has shown that an unauthorized user ("opponent") having only limited material and information with which to work can readily extract the original text of messages enciphered by members of this class by making use of the speed, capacity, and computational abilities of the computer. Representatives of this cipher-class have historically been kept distinct under the names, "Vigenere cipher" and "Vernam cipher," but in spite of superficial differences, their underlying mathematical structures are fundamentally alike. Tuckerman therefore employs the phrase "V-V systems" to encompass them both. In its simplest form, the Vigenere cipher is familiar to even the amateur cryptologist as the "tableau" of alphabets shown in figure 1 and 2. The horizontal and vertical alphabets set slightly apart from the main block are for the convenience of the user in specifying the characters Encipherment of "classic Vigenere" begins by establishing a one-to-one correspondence between the characters of the plaintext and the characters of the key (see figure 1), the key being partially or completely repeated if shorter than the plaintext. The cipher character may then be defined as the character standing at the intersection of a plaintext column and the appropriate keyletter row. If, for example, "EXIT" is used as the key to encipher the plaintext message, "COME HERE," the tableau provides a ciphertext of GLUXLBZX: G at the intersection of "c" column and "E" row; L at the intersection of "0" and "X"; and so on. (Most encipherments omit spaces between words and sentences, since they can provide valuable hints to the cryptanalyst-the code-breaker. However, a good enciphering system should not need this precaution.) There is an extremely simple rna thema tical model which can represent such encipherment. If the letters of the alphabet are assigned numerical values A=O, B=I, . . . Z=25; if combination of a column and a row is defined as "addition"; and if the character at the intersection is defined a~ the "sum" -then encipherment consists of addition modulo 26. According to such addition, 21 + 7 = 2. Inspection of an appropriately labeled tableau (figure 2) shows that the corresponding "letter equation" also holds true: V + H = C. Under this modu- '. COMPUTERS and AUTOMATION for April, 1972 13 1) Ciphertext BMZX ZJ KCVFMEGRI EWDOCX UCNWZ ZSKBLNTFZ FPNSKKXGRFDTKVKVRNSB TVSJONTDWPFY I ZJ NDOGIDNKI XOVZHMDNSEVONUNDNTG YEVFCJ PZPFH PDGJ NGDOOPFDNTK E SPL VYTKRR NPJDFFGRXEETTWMLFMCZ Z VFPPQYZEMDYDEH YRSUJ FDCUZIC ZJ Y BUODDJ PE FUZEEB LNTFZKHNJ XII c[] Q o~o ~. c:. ~ 0 • .. fA an O~"<>-1' /",,'0 """"-.0 · o·0 ~ 2:~ 1C 0 Q ....... r:::=:::J. ' Z .6 0 ~'-:: I .1. 'J ' , ~,,- .......... ........ ...... £! o 0 P 4 0 0 -.~ • c:J' 00 N H 0 '7 8 X F 0 00 (0 0 0 0 . • • ·· V"v .. 0 • · D 5 V D B 0 a lito I ~l ;mp~:11 ;;:" lit8T~ I ~ fJ -=#= [p 00 ~ 0 <1 9[ Item D C E F H I Item C D E H I Answer: J F G J G 1 In each cell, 6 2 insert the 7 code for your 3 8 4 choice of 9 answer: 5 10 Survey Data: 1. Name 2. Title 3. Organization,______________________________________________________________________________________ 4. Address Good? Other (121ease Sl2 ecif y) Excellent? Not your field? Average? 5. In camp uter programming, are you: 6. In syst ems analysis, are you: 7. In mana ging, are you: 8. What fields (not mentioned above) are you fairly good In (or even expert In)?_--,--________________________ 9. What other capacities do you have? (Please don't be bashful -- but be objective) ________________________ 10. Any remarks?________________________________________________________________~~~------~----~~ (attach pal2er if needed) When completed, please send to: Neil Macdonald, Survey Editor, Computers and Automation, 815 Washington Sf., Newtonville, MA 02160 COMPUTERS and AUTOMATION for April, 1972 27 '* * '* * • DO YOU LIKE COMPUTERS AND AUTOMA TION? • DO YOU HAVE A FRI END WHO MIGHT LIKE IT? • HOW ABOUT GIVING HIM (OR YOURSELF) A GIFT SUBSCRIPTION? A reprint from CDrnl?n~a~~J:~ FOUR-5T AR REPRINT-----. SCIENCE AND THE ADVANCED SOCIETY, by C. P. Snow, Ministry of Technology, London, England (April, 1966) THE INFORMATION REVOLUTION AND THE BILL OF RIGHTS, by Dr. Jerome B. Wiesner, M. I. T. (May, 1971) EMPLOYMENT, EDUCATION, AND THE INDUSTRIAL SYSTEM, by Prof. John Kenneth Galbraith, Harvard Univ. (Aug. 1965) COMPUTERS AND THE CONSUMER, by Ralph Nader, Washington, D. C. (Oct. 1970) cut here and tuck in flap r--------------------------, To: Computers and Automation I I I I • Take advantage of our current offer: a gift subscription to him (or you) and for you a copy of our FOUR-STAR ~:,~:,~:,~:, REPRINT Name Address I I ( I!! I evl -= I ANY MORE FRIENDS OR ASSOCIATES WHO MIGHT LIKE SAMPLE COPIES OF CGA FREE? - just tell us and we will send them a copy - so they also may start liking eGA B: I : I I I I I I u I State Zip \ Send 13issues including directory, $18.50 II in U.S.; add 50 cents for Canada, $3.50for Ig I~ foreign postage I~ Send 12 issues without directory, $9.50 in I; U.S.; add 50 cents for Canada, $3.50 for \ foreign postage I enclose $ in full payment. \ ) : I I I I I : Ci ty II • Please enter a subscription for ) I I I Please send a FOUR STAR ~:,~:,~:,~:, REPRINT to me: \ I I I Name_______________________________ : Address \ : City _ _ _ _ _ _ _ _ _ _State _ _ _ _Zip_ _ _ _ _ ; cut here ... M ~ e:{ ..J U t- en a: u:: M ci Z ~ ~ ~ · Ii '"C ·~ III Z· 'J: ot- :1l en 0 a: co w U start my subscription to COMPUTERS AND AUTOMATION according to the instructions checked below. 4/72 One Year (including the Computer Directory and Buyers' Guide - 13 issues) U.S. only. 0 $18.50 One Year (excluding the Computer Directory and Buyers' Guide - 12 issues) U.S. only. 0 $9.50 Z "C ~ 'c g :::> c: ~ YES, '"C • c: e:{ ~ _ _ _ _ _ _ _ _ _ _ _ Fold here - do not cut ::E co ~ __ _ eli '" ~ .z: __ __ _ ~ en e:{ U5 II _ '"C 0- ~ __ _ ev III I co ~ ~ > ..J 0- w a: en en w l- >w ...J w ~ CD W Q.. « en W en +" E ..- C E w Z ~ 2b Ci5 :J co 00. ~ 0 Z 0 0 2 0 0. ~ +" >CD ~ « 0 I::::> :;{ CD a: ...J ...J I- ~ w (!) <{ I- en 0 Q.. W ::::> Q) Q) ~ c: 0 en r.: J: en C'a ~ 0 I.t) t- U CD N 0 co o Payment enclosed fJ Bill me « ~ .!E." .:;: c: 0 +" ~ Q) 2 o o CO Renewal g New subscription ... J: ~ ::I c: III III To SPEED the processing of your order, please check the one best descriptor in each of the two categories below. (This information will be used for statistical purposes only.) BUSINESS TYPE +" 0... ~ ~ Country if not U.S.: _ _ _ _ _ _ _ _ _ _ __ Signature: _ _ _ _ _ _ _ _ _ __ P.O. No.: _ __ « "C ~ ev ev City: _ _ _ _ _ _ _ _ _ State: _ _ _ _ _ Zip: _ _ __ 2 W a: u Address: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ 0 c ~ Name: _ _ _ _ _ _ _ _ _ __ Title:._______ Organization : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ IZ -5 'ro ..J (Ii w en a: Q.. a: 01-Computer Manufacturer 02-Aerospace ,I Ail craft Manufacturer 03-0ther Manufacturing 04-Raw Materials Processing; (chemical, primary metal, petroleum, food. etc.! 05-Mining and Construction 06-Computing & Consulting 07-Finance, Insurance. Pub!., -and Service Organizations DB-Transportation Companies 09-Public Utilities 10-Research II-Wholesale. Retail. Sales, and Marketing Firms 12-Educational; (College, University, or School) 13-Government and Military 14-libraries JOB FUNCTION ___ I-Technical Management; (computer installation management, program management, or engineering mgmU 2-Computer Center Personnel; (methods & procedure analysts. and operators) 3-Programming Personnel; (system5. application & research programmers) 4-Professionai: (systems analysts. mathematicians. operations researchers. and professors) 5-General Management Executives; (corporate officers. owners. and partners) 6-ERgineering Personnel; (systems engineers. research & development engineers) 7-Research Personnel B-Students 9-library Subscription 10-Subscription in Company Name Only I e:{ ~ ev ... U U The Bad Image That Computers Are Earning "Shall we as computer professionals correct our programs, so that they function properly, or wait until we are kicked in the teeth by an enraged public?" a 1. From Harold W. G. Gearing Oueenswood House Stone Drive Colwall Malvern, Worcester England Many of us here are concerned with the bad image that computers are getting throughout the world, from bad data, bad programming, and inadequate training of data-preparation staff. I would suggest that you might open a corner of your magazine to hold bad examples. In your June issue, you commented: "Shall we as computer professionals correct our programs, so that they function properly, or wait until we are kicked in the teeth by an enraged public?" Here are a few to be going on with: 1. A well-known British manufacturer of motor-car accessories put its accounts-payable onto a computer. The young girls punching the input cards from the supplier invoices were not properly trained, and did not appreciate how to distinguish goods value from returnable outer-packaging deposits. One supplier found that over a period of several months, items in the Company's details of payment forms bore little relationship to his statements for accounts receivable. A day trip of several hundred miles was necessary, to persuade the responsible accountant to examine his internal system. g 2. A security-transport company computerised its sales invoices. They show only contract number and no narrative. Central offices of large users have to examine contracts each time invoice batches arrive and manually allocate the invoices by writing on them the source and destination of each service. The invoicing system appears to be geared to holding a weekly service charge only; hence when the service is rendered only monthly, the contract monthly charge is divided by 4.33, rounded for entry in the master record; subsequent multiplication by 4.33 by the standard monthly invoicing program results in an invoice which differs by pennies from the contractual figure, which in many cases is also being prepaid by a banker's standing order. 3. A London west-end outfitting company decided to save money on its sales-ledger system by notifying all its account holders that they were being transferred to Barclay-card. Many English people dislike credit cards, and future business is probably being lost, particularly postal order business. In this country the Inland Revenue is modifying its former ambitions to centralize all income tax records in a few computer centres. British obstiCOMPUTERS and AUTOMATION for April, 1972 nacy, in continuing to prefer to deal with a local tax inspector, may be winning the day against the Big-Brother Computer! 2. From The South Middlesex News, Jan. 18, 1972 Framingham, Mass. In this age of mushrooming automation, man has scored a rare victory over a computer -- in this case a $1.5 million machine described as a "good worker" which "just couldn't compete with people." In Sacramento, Calif., in a move that cut teacher credential processing time by 900 per cent, George Gustafson fired the machine and switched to human beings. "We got rid of a million and a half dollars worth of computer," said Gustafson, executive secretary of the State Teacher Preparation and Licensing Commission. "We pulled the plug and sent it back to IBM." Gustafson Monday described his action as "converting credential processing from a complex and costly automated system to a streamlined, fully manual operation." "It was easier to do it by harLd," said Gustafson, who gave this assessment of the departed computer: "It was a good worker; it jus t couldn't compete wi th people." This giant step backwards, Gustafson said, resulted in cutting the credential period from an average of 95 days down to the current average of 10. Gustafson said that as a result of the change he was able to reduce the staff from 240 persons to 106. Since August, the department has cut 80 employees but all but three were given jobs with other state agencies. A side benefit of the switch came during the summer, when the commission hired more than 50 poor and minority students to help make the change to manual processing. ("Man Beats the Machine", by United Press International) 3. From New York Daily News, Jan. 23, 1972 New York, N. Y. More than 100 state senators and assemblymen joined Sen. John D. Caemmerer (R-Nassau) yesterday in sponsoring two bills to curb the power of the Parking Violations Bureau, which has harassed and dunned thousands of innocent motorists to pay for undeserved tickets. Caemmerer, chairman of the Joint Legislative Committee on Transportation, said passage of the 29 bills would "prevent another PVB nightmare New York City-or any other city." in The antics of the bureau and its computer, Super Sleuth, were detailed in a News series after thousands of readers pleaded for help by writing Action Line, a reader service feature that runs in the Long Island edition. Motorists who never had driven in New York City were being plagued by Super Sleuth to pay parking tickets. Letters to the bureau, at 475 Park Ave., South, brought no reply-only higher fines and stepped-up threats. The protests resulted in public hearings by Caemmerer Dec. 2 at the State Office Building, at which exasperated motorists told of widespread bureaucratic bungling and arrogant disinterest by the bureau. Caemmerer's bills would require the violations bureau to provide transcripts of all hearings, requ~re police witnesses if requested and allow defendants to be present at app~als. Bureau Director Anthony Atlas has often declared that the abuses of the bureau were being corrected, but letters to The News from harassed and distressed motoris~did not diminish. "My mail is worse than ever," Caemmerer said. At The News, the rate of complaints also is higher than ever. ("Aim Monkeywrench at Super Sleuth", by Donald Weinbrenner) 4. From The New York Times, Feb. 20, 1972 New York, N. Y. In the two weeks since 155,000 maximum base rent orders have gone out to tenants living in rentcontrolled apartments, 17,379 landlords and tenants have visited or telephoned the city Office of Rent Control to protest or challenge the new rents fixed in the orders. Most important-at least to the innocent who are being victimized-the bills would require more careful vehicle identification and would require the bureau to answer its mail. One was an owner who had carefully reported on the proper form when he increased the rent on an apartment that had become vacated, only to have it ignored when the city's computer figured the maximum rent he could collect from it. "We must deal with the PVB's preposterous attitude toward answering legitimate mailed-in questions," Caemmerer said. Robert C. Rosenberg, the city official in charge of the office's computer, traced the error and partly calmed the landlord. The most maddening situation facing many motorists was that a parking ticket would be sent to them in Nassau County, although they had never driven in New York City. Not only that, it was marked "second notice." It turned out that the owner had listed the apartment as "45" on one form and "4-5" on another. Computer Confused In a typical example, a man with a blue Rambler, license plate XYZ-1234, would get a "second notice" to pay a ticket issued in the Bronx-where he had never been-on a yellow Ford with a different plate number. Mr. Rosenberg, assistant administrator of the Housing and Development Administration's maximum base rent system, said his computer had a onetrack mind and did not realize that "45" was "4-5" as well as-on some occasions in some landlords' minds-"S4. " He then would write a letter-registered-to explain the goof. A month later, he would get a more threatening letter demanding a higher fine. This would continue implacably, willi higher fines and threats to withhold auto registration renewal. "The computer isn't able to match the listings," Mr. Rosenberg said. "A special uni t.-of people, not computer sections-is resolving these matching problems. As soon as a problem is resolved, the computer is notified and a corrected order issued." So the motorist was forced to pay up-although innocent-or take a day off from work and drive 50 or 60 miles to straighten it out. In many cases, the motorist would drive in, be told: "Don't worry about it," and the next week receive another threatening letter. Mr. Rosenberg said landlords and tenants who discovered apparent errors should file protests at once. A tally Friday, however, showed that the 17,379 persons, who own or rent 21,250 apartments, did not wait for urging and have filed protests already. More will be coming in next week, for 102,000 orders are to go out tomorrow. Caemmerer's bill would require the cop or meter maid to put more information on the parking ticket. In addition to the make of the vehicle, the bill would call for it,s color, model, body type, plate number and registration tag number, with its expiration date, accordnig to the Senator. This should help the Motor Vehicle Department in Albany identify more carefully the owner of the car and make sure the ticket reaches the person who deserved it. In all, more than 1.1 million orders are to go out. One set of problems that is giving the computer pause arises from inconsistent financial data. When data are not consistent, the computer does not attempt to reconcile the conflict. It just omits the apartments to which they refer. The legislation also would require that all certified mail tothe bureau from motorists must be answered wi thin 30 days or the case must be dismissed. Until it gets a correction, the computer solves its problems by tapping out a message on the order: "Controlled apartments for which you have not been notified require further research. You will receive notification on these apartments shortly." The buteau had never bothered to answer mail. Letters were stacked up by the thousands. ("17,379 Protest New Rent Orders; Inconsistent Form Entries and Computer Blamed", by Will Lissner) 30 II COMPUTERS and AUTOMATION for April, 1972 .~ Do You Want To Stop Cri.me? William P. Wood, III Dept. of Information Systems Virginia Commonwealth Univ. 800 W. Franklin St. Richmond, Va. 23220 "The last time the three fugitives were seen after blowing up the University laboratory, they were driving north towards Canada." '~he robbers jumped into a waiting car and fled into the night." Automobiles Used in Crime Such are typical daily news reports. What do all of these reports have in common? Almost every criminal MUST USE AN AUTOMOBILE SHORTLY BEFORE AND/OR AFTER HIS CRIME. Automatic Car Identification Such a fact might be used to great advantage in law enforcement. A system that would utilize this fact is as follows: The principle used is borrowed from the railroad industry. The railroads use a system named ACI (Automatic Car Identification) to daily track and locate railroad cars. A problem the railroads have had until ACI is the losing of railroad cars. It is not unusual for a boxcar to be hooked up to the wrong train, sent across the country, and sidetracked on some remote spur. To remedy this problem, the railroads set up a system of photoelectric sensing devices along their tracks. These devices have the ability to read color coded plaques placed on the side of the cars. These sensing stations are connected, via telecommunications, to a central computer complex. This complex stores the location of the industry's two million cars. License Plates Read by Sensing Stations Using a similar system the nation's automobiles could be kept track of for the purpose of law enforcement. License plates would be required by law to conform to specifications which would allow them to be read by the sensing stations. These stations would be set up at strategic locations along our nation's highways. They would be tied via telecommunications equipment to centrally located computers manned by law enforcement personnel. Suggested Systems to Track Criminals This system could be constructed for either of two levels of capability. The more limited level would be that used for only tracking specific vehicles. For example, this would track known criminals. The more capable, and thus more expensive system, would sense and keep in memory for short periods, perhaps a few days, the location of all vehicles on U. S. roads with license plates. Information concerning cars of special interest could be kept in memory as long as necessary. Such a system would provide the following advantages: 1. All cars within the area of a crime before and immediately after could be identified and their license numbers checked against suspect's license numbers. COMPUTERS and AUTOMATION for April, 1972 2. Getaway cars could be traced throughout the country. 3. Suspects and witnesses would have their stories verified or discounted. 4. Such a system would be a great deterrent to crime. 5. Stolen cars could be traced almost immediately. 6. Drug traffic could be traced. 7. Such a system would cut into virtually every kind of crime: murder, robbery, rape, drug, arson, kidnappings, bombings, vandalism. .Cost of Tracking System and Cost of Crime The cost of such a system might climb above one billion dollars, just for the setup cost. The cost of crime, though, is well in excess of twenty-one billion dollars per year. Even if the system did not pay for itself in the dollars and cents terms of cutting the cost of crime by as much as its own cost, how can one put a price tag on the misery and fear generated by this plague on society? Citizens, fearing for the safety of their families and homes, flee to the relative safety of the suburbs. Parents pray that their children will not fall prey to the evils of drugs and juvenile delinquency. Not only would this system be a boon to rid our society of those sinister doers of evil, but it would act as a tremendous before-the-act deterrent to those realizing the probability of apprehension. Cost to Our Personal Privacy There is one final cost that remains to be considered. This is the cost to our personal privacy generated by the ever encroaching inroads of computers and electronic gadgetry. Would such a system shorten the time between now and 1984? IBM'S POWERFUL PARTNER: The Accounting Principles Board (Editorial, reprinted with permission from "Samson Technology Trends", Dec. 1971, copyright 1971 by and published by Quantum Science Corporation, 851 Welch Road, Palo Alto, Calif. 94304) More than the increased competition from IBM, the recent ruling from the Accounting Principles Board (APB) , threatens to decima te the ranks of independent computer companies that supply plug-compatible peripherals to the end user. The high principled accounting rules are forcing all companies, large or small, to suddenly be super-honest and as a result, the APB has forced the majority of the small and medium sized independent computer companies to their knees, and sometimes into bankruptcy. The ruling, instituted last year, requires that a conditional sale to a third party lessor be treated on a standard rental accounting basis rather than as a sale. This ruling has drastically reduced the reported earning of all peripheral suppliers who lease the majority of their equipment and has made the financing of rental inventory almost totally impossible for most. In fact, no company which does not have a sizeable two to four year-old base of rental equip- 31 ment in the field can report profit. Worst of all, the fastest growing companies are penalized most. Before this ruling, most peripheral equipment suppliers relied on third party leasing to finance their sales. Moreover, the third party transactions provided the independents with an excellent profit record as long as their sales base continued to grow. Third party sales were recorded as revenue at the full sales price of the leased equipment rather than only the rental revenue accruing to the leasing company. Admittedly, the old treatment unfairly favored the peripheral equipment supplier because it did not fully allow for the potential ri sks and wri te-offs resulting from the return of obsolete equipment. It was nevertheless approved of and certified by the accounting profession for several years. The ethics of such misguided support of a false business practice must seriously be questioned, as well as the sudden application of new principles of accounting that reversed prior certified profits into losses. The APB, in an attempt to place the profi tabili ty of the independents in proper perspective, has now created a situation that makes sales growth of computer products to the end user nearly impossible except for large companies that can finance their own leases. Many small companies that based their business plans on old accounting rules have already or will soon cease existence as independent entities. The APB now has on its conscience not only dozens of bankruptcies, but also large losses of hundreds of thousands of stockholders who had accepted their views and relied on them to make investment decisions. What can we look forward to from the accounting profession in 1972 after the pooling-of-assets ruling in 1970 and third party lease accounting in 1971? Happy New Year! CDC VS IBM - CORRECTION /. From Frederic O. Par/ova 169 West 88 St. New York, N. Y. 10024 I read with interest the article on the IBM/ CDC/Greyhbund Litigation (C&A, Feb. '72). In your introductory remarks you state, " ... and there is apparently nothing in the court order(s) which enables the plaintiff, CDC, to obtain access to the information ...... " May I call your attention to page 49? "Examination and evaluation by expert IBM, Greyhound, and Control Data personnel is necessary if the data is to have any meaningful significance to counsel in preparing their defense." Also, "IBM shall make the said answers available only to its counsel and to not more than 15 of its non-clerical personnel .... CDC and Greyhound shall comply with this paragraph but with a limitation of 10 non-clerical employees .... " II. From the Editor We regret the error, and appreciate very much your correction. CORRECTION In the March 1972 issue of Computers and Automation, the following correction should be made: page 27, "Pictorial Reasoning Test: C&A No.5" - in the first paragraph under Figure 2, replace the words "include OK or X as appropriate" by "include V, X, or Z as appropriate." 32 REDUCING AND DISMANTLING SCIENCE AND RESEARCH INSTITUTIONS, AND SOCIAL RESPONSIBILITY Andrew G. Michalitsanos University of Cambridge Cambridge, Eng/and The disillusionment with science in the United States is clearly evident in the continuing decline in the number of private and industrial research positions and the systematic reduction of research funds in both governmental and nongovernmental sectors. The attitude of the Nixon Administration and the U. S. Congress has been aimed at drastically reducing the volume of scientific research in the United States, a nation whose growth and prosperity is almost totally dependent on the inventiveness of its scientists, who can provide new ideas and techniques for industry. The Government's reason for this policy is based on public clamor against "non-relevant science." It would be interesting indeed to find a person who could define" relevant scientific research." Many areas of scientific endeavor were at one time or another considered not directly applicable to public needs, but have become almost guiding centers of industrial applicability many years later. For example, solid state physics, at one time only of academic interest, today occupies an important place in electronic applications of all types. The telephone on the President's desk was constructed with the aid of "non-relevant scientific research." The U. S. Government's attitude has a much deeper psychological effect on the American public in general. The cutting of research funds fosters the belief held by many young citizens today that science lies at the roots of the country's dilemma since it has brought the problems as well as the advantages. The new generation appears to have replaced scientific logic with mysticism, clearly a step backward into the caves. Pure science has not created the problems, but rather society's inability to apply scientific discoveries in a resourceful and meaningful manner. We should recognize that pure scientific research provides an intellectual store from which we may extract ideas and apply them to our needs. Reducing funds and decreasing the amount of pure scientific rese~rch simply makes the available store of informa tion smaller. I am not advocating indiscriminate spending for relatively obscure areas of science. However, I can see little use in dismantling the research institutions that have taken decades to establish. The Nixon Administration should reassess its actions and consider what sad effects their decisions will have on the United States and the rest of the world for decades to come. (Based on a letter published in the New York Times, Sept. 24, 1971) COMPUTERS and AUTOMATION for April, 1972 •• A Concerted Campaign To Deny The ~merican People Essential Knowledge About The Operation of Their Govern ment Prof. Henry Steele Commager Amherst College Amherst, Mass. 01002 The Nixon Administration is at it again. First it tried to intimidate the television networks. In vain. Next it attempt~d, for the first time in our history, to silence the press by the threat of prior censorship. Again in vain. Now on the branch United in the the Justice Department has launched an attack constitutional privileges of a co-ordinate of the Government -- the Congress of the States and, by implication, every legislature United States. The issue is once again the Pentagon Papers. We might have supposed that with its defeat in The New York Times case the Government would drop thi s shabby prosecution. Not at all. At the instigation of the internal security division of the Justice Department, a Federal grand jury in Boston has now subpoenaed Dr. Rodberg, legislative assistant to Senator Gravel, to appear before it. The Government's brief acknowledges that the thrust of the subpoena is to prepare the ground for an inquiry into "acts done by Senator Gravel in reading and inserting into the record the Pentagon Papers." This may seem like a tempest in a teacup -- especially as the Government has by its decision to publish the Pentagon Papers in toto abandoned its argument that their publication would do "irreparable injury" to the "security" of the United States -an argument palpably absurd at the time. But in fact three major principles of our constitutional system are at stake in this new Justice Department caper. First is the hard-won principle of the immunity of any legislator from just this sort of harassment. Second is the principle of the separation and the equality of powers of the three departments of government. Thi rd is the principle of freedom of speech and of the press. , The Constitution provides (Art. I, sec. 6) that "for any speech or debate in ei ther House, they [Congressmen] shall not be questioned in any other place." Neither the purpose nor the meaning of this clause is obscure. The purpose was to make it forever impossible for any executive authority to punish or intimidate any legislator for what he might say in the legislative chambers. These interpretations are now challe~ged by the Government brief which asserts that Congressional immunity from "question or debate" does not cover committee reports, and that it is wholly personal and cannot be enlarged to embrace legislative staff. The first of these assertions was disposed of as recently as 1969 when Chief Justice Warren reiterated the principle that the speech clause covered committee reports, resolutions, "and such things as are generally done in a session of the House in relation to the business before it." COMPUTERS and AUTOMATION for April, 1972 The second assertion raises the larger issue of the separation and equality of powers in our Government. But the principle laid down in Jefferson's Manual of Parliamentary Practice and formally adopted by both houses of Congress, has never been successfully challenged, name ly tha t "Congressmen are at all times exempted from question elsewhere, for anything said in their own House; that during the time of privilege neither a member himself, nor his wife, nor his servants, may be arrested, cited, or subpoenaed in any court." The logic of these constitutional and legal immunities is obvious. It is not to confer special privileges on legislators -- or on judges or Presidents -- but to protect them against harassment and intimidation which, as Justice Harlan put it, might "inhibi t and dampen the ardor of all but the most resolute, or the most irresponsible, in the unflinching discharge of their duties." The attack on Senator Gravel's immunity is pernicious -- not only constitutionally and politically, but philosophically. It is part and parcel of what can only be described as a concerted campaign to deny the American people that knowledge about the operation of their Government so essential to the sound functioning of democracy. It is a direct assault on the Constitution and the separation of powers; it is an indirect assault on the principles which the Constitution was designed to preserve and advance, above all the principle of freedom of speech and of the press. In this matter what the father of the Constitution, James Madison, said in 1794, is still relevant: "If we advert to the nature of Republican government, we shall find that the censorial power is in the people over the I Government, not in the Government over the people." (Based on a report "A Senator's Immunity" by Henry Steele Commager published in The New York Times, October 15, 1971) FAll JOINT COMPUTER CONFERENCE: TOPICS From the Editor The Fall Joint Computer Conference in its call for papers has listed the topics which it seeks or hopes to cover in the program of papers: User Applications and Requirements Architecture Trends and Limitations Terminals Communications Complete Systems and Networks Hardware Advances Software Development Measurements, Analysis, and Evaluation Reliabili ty Social Issues New Ideas -- Try Us Some of the most important topics in the computer field are left out of this list: Computers and Privacy Should IBM be Broken Up as a Monopoly? The Distrust of Computerized Systems by the Public The Use of Computers in Dictatorships Should Computer People Develop into Information Engineers? Perhaps all these topics arc included silently in the topic "Social Issues". 33 Internal Revenue Service: USE OF COMPUTERS William H. Stewart, Jr. Asst. Prof. of Political Science Bureau of Public Administration Univ.. of Alabama University, Alabama (Excerpt from "Computers and Government", Citizens Information Report No.7, publi shed by Bureau of Public Administration, University of Alabama, 1972, 54 pp. "The material in this publication is not copyrighted in the hope that its use without restriction will be more widespread.") Perhaps the most comprehensive use of computers in American government at any level is that made by the federal Internal Revenue Service. Automatic data processing in this agency has resulted in more accurate identification of potential and delinquent taxpayers, prevention of duplicate refunding of taxes, verification of tax computation, billing of outstanding taxes, tax audit, development of statistical information for proposed legislation, and tax administration controls. These improvements are not simply to the government's benefi t. They help ensure that the tax burden will be borne equitably and that some do not escape paying legally owed taxes while others pay more than they are obliged to pay. Approximately 76.8 million individual income tax returns were filed during the 1970 filing period. The role of computers in the processing of these returns is substantial. The central automatic data processing installation of the Internal Revenue Service is located at Martinsburg, West Virginia. The system is so massive that it requires 275 technicians to feed in the data received from district computer centers, which handle initial processing and verification of tax returns. Returns are transmitted to Martinsburg for more extensive analysis and consolidation into the master magnetic tape files. One-half inch of tape contains a complete three-year running tax account for each return filer. The automatic data processing system utilized by the Internal Revenue Service has been programmed to look for the most common types of taxpayer "mistakes" and the classifications of taxpayers who are most likely to make these errors. The agency's experience is that the higher the individual or corporate income the greater is the chance of error. Its data system program includes a special check for returns over a certain secret figure. The computer will note and flag returns which depart markedly from expectations. An audit may be indicated for taxpayers who list substantially more in the areas of charitable contributions, interest payments, and medical expenses than they have in previous years and/or exceed the normal figures expected for individuals in their income bracket. A combination of incongruencies will increase greatly the likelihood of an audit by mail or through a personal interview. At the present time approximately one in44 returns is audited. This the computer does not do. It verifies arithmetic, prepares tax bills, and indicates which returns should be audited. The auditing process itself, however, is not automated. Based onthe number of revenue agents it has, a district Internal Revenue Service office advises the Martinsburg computer system of the number of return~ it has the manpower to audi t. Given a figure of, for example, 75,000, the data processing center supplies the office with the names of the 75,000 "best" candidates for 34 further examination based on how far these people depart from the standard categories. Federal officials feel that automatic data processing is geared to improving accuracy and equity in tax collecting. The efficiency of the system makes possible the collection of greater revenue at smaller cost. During the initial year in which computerized tax collection procedures were in operation in the Southeast, many tax delinquents paid taxes owed without governmental prodding in the belief that these omissions would be discovered by the government's computers. Computers have had, therefore, not only a technological but a psychological effect on tax collection procedures. Officials of the Internal Revenue Service also report that embezzling by agency employees has become harder since the processing of income tax returns was made the assignment of giant computers. SOME RESPONSIBILITY FOR OUR CHAOTIC SOCIETY S. R. Harrison, Ph.D. The Norwegian Institute of Technology Trondheim, Norway Good for you and your articles on President Kennedy's assassination. Our country needs people like you and your staff who are willing to take a stand and fight for what they believe to be true. The time is long overdue that our profession had some life injected into it and begins to accept some responsibility for our chaotic society. I find too many highly specialized journals "doing their jobs" in presenting only restrictive technical publications. I applaud your efforts. Unsettling, Disturbing, Critical Computers and Automation, established 1951 and therefore the oldest magazine in the field of computers and data processing, believes that the profession of information engineer includes not only competence in handling information using computers and other means, but also a broad responsibility, in a professional and engineering sense, for: The reliability and social significance of pertinent input data; The social value and truth of the output results. In the same way, a bridge engineer takes a professional responsibility for the reliability and significance of the data he uses, and the safety and efficiency of the bridge he builds, for human beings to risk their lives on. Accordingly, Computers and Automation publishes from time to time articles and other informatjon related to socially useful input and output of data systems in a broad sense. To this end we seek to publish what is unsettling, disturbing, critical -- but productive of thought and an improved and safer "house" for all humanity, an earth in which our children and later generations may have a future, instead of facing extinction. The professional information engineer needs to relate his engineering to the most important and most serious problems in the world tOday: war, nuclear weapons, pollution, the population explosion, and many more. COMPUTERS and AUTOMATION for April, 1972 RTR GGO NFH WSW pez Qeo GVR "fF'D [J
Source Exif Data:File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.3 Linearized : No XMP Toolkit : Adobe XMP Core 4.2.1-c043 52.372728, 2009/01/18-15:56:37 Producer : Adobe Acrobat 9.1 Paper Capture Plug-in Modify Date : 2009:03:24 23:25:47-07:00 Create Date : 2009:03:24 23:25:47-07:00 Metadata Date : 2009:03:24 23:25:47-07:00 Format : application/pdf Document ID : uuid:00ea68d2-6650-4d18-b8b1-88f5eb4f08ef Instance ID : uuid:3a9c714e-9e10-4dcb-8b53-065cbda62315 Page Layout : SinglePage Page Mode : UseNone Page Count : 52EXIF Metadata provided by EXIF.tools