Page images
PDF
EPUB

scanners.

I suspect that the development of a scanner that can read hadwriting on a letter is so difficult that it may not yet be available. But the scanner is so important to the economics of storing and retrieving mass amounts of information that there is a tremendous impetus to develop and perfect it. We cannot key punch the Library of Congress into a data bank. The only way we can get the Library of Congress computerized in an efficient, economic way is by means of scanning techniques for input. I would be very surprised if the military and certain other Government groups didn't have optical

Senator Long. Do you know whether there is a device than can be used in an envelope that can be put in under the flap without breaking it open and scan inside the letter?

Professor MILLER. That sounds like something my dentist would use. The only internal scanner I have heard anything about is the one that supposedly reads through the envelope. I honestly do not know anything about an insertable lens.

I should also add that we should not ignore the possible development of techniques and equipment for picking up computer radiations. I have yet to see a definitive statement as to whether or not a computer radiates when it is operating. Quite possibly someone can develop a machine that acts as a supersophisticated hearing device, and can reconstruct the data by recording and interpreting the computer radiations that may exist in the environment of the machine.

Senator Long. You mentioned a while ago that it would be easy to break the code now used-however you described it-but you thought in perhaps 5 or 6 years technology might develop to such an extent that that could not be done. But, if that develops, on the one hand, isn't it typical that it develops on the other?

Professor MILLER. Absolutely. This is going to be a continuing problem. Congress should not establish a data bank and legislatively list safeguards without also creating a mechanism for constant surveillance of the center's operations and updating of the machinery and procedures used by the censor.

Senator Long. You described this as Parkinson's law. I believe it was Judge Brandeis who said that this wiretapping, and so on, was a dirty business. It is more like Parkinson's disease. It is sort of a cancer that eats on the society.

Mr. Waters.
Mr. WATERS. Thank you, Mr. Chairman.

Professor Miller, you have prepared such an excellent statement I wonder if I might impose on you to apprise the subcommittee with respect to your inquiry in the optical scanning field, and if I may ask leave of the Chair to have the record left open so that you can forward for inclusion in the record such information as you might have available that you feel would be helpful to the subcommittee in connection with the types of scanner that you have described.

Mr. MILLER. I have a certain feeling that after this morning any lines of information formerly open to me on that subject may be closed. But I shall try, Mr. Waters.

(The information referred to follows:)

THE UNIVERSITY OF MICHIGAN LAW School,

Ann Arbor, Mich., March 28, 1967. Hon. EDWARD V. LONG, Subcommittee on Administrative Practice and Procedure, Committee on the Judiciary, U.S. Senate, Washington, D.C.

DEAR SENATOR Long: During my testimony before the Subcommittee on March 16 I made reference to developments in the technology of optical scanning and the possibility of these devices being used in connection with mail-cover operations. This letter is in response to Mr. Waters' request to provide the Subcommittee with any additional information on this subject that might come to my attention. trust it will be inserted at the appropriate place in the record.

Optical scanners have been under development for approximately 15 years. By and large, the promises in this area of computer technology have outstripped the commercial realities because of technological obstacles and high costs. The objective has been to develop a device with multi-font capacity, that will operate at high speeds, read reliably (i.e., with a high incidence of character recognition and a low incidence of recognition error), and be economical.

Single-font readers have been available and in use for several years and currently are in operation in several agencies of the federal government. To my knowledge, they can be found in the Social Security Administration, the Air Force, and the Post Office. I am sure that other governmental organizations use them. The largest and most successful scanner manufacturer appears to be the Rabinow Corporation, which is located in Florida. I am told that they have installed about 60 single-font scanners and that these units cost approximately $84,000 each. Scanners also are manufactured by IBM and the Intelligent Machines Corporation.

Multi-font readers have been under development for about a decade and are now being used on an experimental basis in various places. The three corporations mentioned in the preceding paragraph appear to be the leaders in the multi-font field. At present, multi-font readers are available on a customized basis but are not being mass produced. I am told that the Rabinow Corporation is working on customized multi-font readers for the Post Office and that an experimental unit is currently in use in the Detroit Post Office for sorting Zip Code numbers. IBM recently announced a “break-through” in the multi-font scanning field and this may herald the production of such units on a commercial basis.

I have been reliably informed that the new multi-font scanners can be programmed to recognize a variety of type fonts and handwriting. The basic problem has been to develop machines that have wide recognition capacities coupled with high speed and low error rate. By and large availability of these characteristics depends upon how much one is willing to pay for the machine, although the technology still is not matured to the point of not requiring some trade-offs among the three main attributes. I am also told that the type of optical scanning for mail-cover operations I described to you during my testimony before the Subcommittee is feasible and that the transfer of the information picked up by the scanner to individualized computer dossiers is simply a question of proper programming. I trust that this letter will be of some aid to you and the Subcommittee. Sincerely yours,

ARTHUR R. MILLER,

Professor of Law. Mr. WATERS. We have heard the saying, Professor Miller, that knowledge is power. Certainly it appears to me that whoever has the key to the knowledge stored up in this vast apparatus is going to have an awful lot of power.

I thank you very much, Professor Miller. Thank you, Mr. Chairman. Professor MILLER. Thank you.

Senator Long. Thank you, Professor. Your statement has been unusually helpful. I am very much impressed with it, and very grateful for your being here today. Professor MILLER. Thank you, sir. Senator Long. Thank you. That concludes the hearing this morning.

The subcommittee will now stand in recess until tomorrow at 10 o'clock in this same room where we will resume hearings at that time.

The subcommittee is in recess.

(Whereupon, at 12:30 p.m., the subcommittee was adjourned, to reconvene tomorrow, Wednesday, March 15, 1967, at 10 a.m.)

COMPUTER PRIVACY

WEDNESDAY, MARCH 15, 1967

U.S. SENATE,
SUBCOMMITTEE ON ADMINISTRATIVE PRACTICE AND PROCEDURE

OF THE COMMITTEE ON THE JUDICIARY, Washington, D.C. The subcommittee met, pursuant to recess, at 10:10 a.m., in room 1318, New Senate Office Building, Senator Edward V. Long, of Missouri (chairman of the subcommittee), presiding.

Present: Senators Long, of Missouri (presiding), and Thurmond.

Also present: Bernard Fensterwald, Jr., Chief Counsel; Bernard J. Waters, Senator Dirksen's Office, Minority Counsel; and Benny L. Kass, Assistant Counsel.

Senator Long. The committee will be in order.

The committee has a telegram from the director of the New York State Identification and Intelligence System, of Albany, dealing with the subject at hand which, without objection, will be placed in the record at this time. (The telegram referred to follows:)

ALBANY, N.Y.,

March 15, 1967. Senator EDWARD V. LONG, Chairman, Subcommittee on Administrative Practice and Procedure, New Senate Office Building, Washington, D.C.

DEAR SENATOR LONG: It has been brought to my attention that professor Arthur Miller, school of law, University of Michigan, in a statement before your subcommittee on March 14, 1967, referred to my testimony on July 28, 1966, concerning the New York State identification and intelligence system before the House of Representatives' Special Subcommittee on Invasion of Privacy. Some of Professor Miller's statements do not fully reflect the import of my testimony in that:

(1) Professor Miller says: "A massive computer installation containing extensive files on criminals and alleged criminals that can be reached simply by digit dialing and is available to governmental or private offices not involved in law enforcement poses a serious threat to the individual. Nowhere in my testimony do I make or imply such a statement. On the contrary, the record of transcript reads as follows:

“Mr. ROSENTHAL. Inspector Gallati, does the statute that set you up limit the people that you can provide information to ?”

“Mr. GALLATI. This sets up the limitation that is designed for qualified agencies concerned with the administration of criminal justice. And it expressly states that this means courts of record, probation departments, sheriffs offices, district attorneys' offices, State division of parole, New York City Parole Commission, State department of correction, New York City, department of correction, and police forces and departments having responsibility for enforcement of the general criminal laws of the State."

(2) Professor Miller states: "A State official may direct an inquiry to the computer file of 'known' criminals, find an entry under the name of his subject, and rely on that entry to the subject's detriment without attempting to verify its accuracy or exploring the possibility that additional information might eliminate the inuendo of the computer record."

101

« PreviousContinue »