American Journal of Computational htguistiet Microfiche 43 
THE FINITE STRING 
NEWSLETTER OF THE ASSOCIATION FOR CORUTATIONAL LINGUISTICS 
The members and Executive Committee of the Association will 
soon have to choose among limiting the scope of the Journal, 
making its content more compact and its content more con- 
cise, or increasing dues. The production budget allows for 
no more than 20 fiches per year. The rate of arrival of ac- 
ceptable manuscripts and news that's fit to print is going 
above the level that can be accommodated in that budget. 
Members who prefer one of the three recognized courses of 
action, or who have another to propose, can write to the 
Editor, the President, or any member of the Editorial Board 
or Executive Committee. 
AMERICAN JOUFWL OF COMPUTATIONAL LINGUISTICS is published 
by the Center for Applied Linguistics for the Association 
for Computational Linguistics. 
EDITOR: David G. Hays Pmfessor af Linguistics, SUNY ~uffalo 
EDITORIAL ASSISTANT: William Benzon 
EDITORIAL ADDRESS: win willows, Wanakah, New York 14075 
MANAGING EDITOR: A. Hood Roberts Deputy Director, Center for 
Applied Linguistics 
MANAGEMENT ASSISTANT : James Megginson 
PRODUCTION AND SUBSCRIPTION ADDRESS: 1611 North Kent Street, 
Arlington, Virginia 22209 
Copyright 0 1976 
Association for Computational Linguistics 
TABLE OF CONTENTS 
PERSONAL NOTES ............ . . . 3 
1976 NATIONAL COMPUTER CONFERENCE ... . . 4 
NIMH: TECHNICAL HELP FOR PROPOSERS . MINORITY PROGRAMS . 5 
NSF: FOREIGN CURRENCY PROGRAM..EGYPT . INDIA . PAKISTAN . 6 
CATASTROPHE THEORY: THOM AT SIAM MEETING . . ... 7 
NATO8 ADVANCED STUDY INSTITUTES ........ 8 
STRUCTURAL-PROCESS THEORIES OF BEHAVIOR .... 8 
MAN-COMPUTER INTERACTION ............ 9 
COMPUTER-BASED SCIENCE INSTRUCTION .... . 10 
C . 5 . PIERCE INTERNATIONAL CONGRESS . . ...... 11 
NEW JOURNAL: COGNITIVE SCIENCE .......... 12 
NSFI REJECTED PROPOSALS G RECONSIDERATION .... 13 
CONFERENCE CHRONICLE ........ ...... 14 
1976 LINGUISTIC INSTITUTE. OSWEGO . NEW YORK .... 15 
BAAL: SEMINAR ON TRANSLATION AT ESSEX. ENGLAND ..... 19 
INDEXING IN PERSPECTIVE SEMINAR AT WARSAW . . 20 
NFAIS OFFICERS 1976-77 ................ 21 
EXPERIMENTS WITH A POWERFUL PARSER (REPRINT) Martin Kay 22 
AFIPS WASHINGTON REPORT (REPRINT. 1975-76 ISSUES) .... 53 
American Journal of Computational Linguistics Microfiche 43 : 3 
PERSONAL NOTES 
JAIME CARBONELL MEMORIAL 
A Japanese translation of REPRESENTATION AND UNDERSTANDING 
is in preparation. dedicated to Carbonell's memory. This 
edition will include a paper signed by Carbonell and 
Collins, published posthumously in AJCL 
BRUDERER, HERBERT New address: Finkenweg 3, 3110 Munsingen, 
Switzerland 
NOVAK. GORDON S . JR. PhD in Computer Science, University of 
Texas, Austin, for a thesis entitled COMPUTER UNDERSTANDING 
OF PHYSICS PROBLEMS STATED IN NATURAL LANGUAGE 
American Journal of Computational Linguistics Microfiche 43 : 4 
1976 NATIONAL COMPUTER CONFERENCE 
NEW YORK HILTON HOTEL AND COLISEUM 
JUNE 7 - 10 
TWENTY-FIFTH ANNIVERSARY. The first Joint Computer Conference 
was held December 10-12, 1951, in Philadelphia. Ten large- 
scale computers were described: UNIVAC, Burroughs, IBM CPC, 
ORDVAC, ERA 1101, MARK 111, Ferranti-Manchester, Whirlwind I, 
EDSAC, NBS SEAC. Bell Labs was in the field. 877 attended. 
KEYNOTE ADDRESS. J. Paul Lyet, Chairman of Sperry Rand Corp. 
PUBLIC ACCESS TO COMPUTERS: Tuesday 8 June. David Ahl, chn. 
Public attitudes toward computers, Personal computers, Access. 
INTERNATIONAL PLENARY SESSION. Monday 7 June. Bob 0. Evans. 
Japan, UK, USSR, Central Europe. 
ART EXHIBIT. Systems Dimension, Ltd., ICCH/2 Touring Exhibi- 
tion, Henry Christiansen Collection, other works. 
INTERACTIVE AND NETWORK DEMONSTRATION. Telenet nationwide 
packet-switching network--first U.S. carrier in the field. 
Several dozen terminals in the Coliseum. Econometric fore- 
casting models; information retrieval; conferencing; editing; 
engineering design graphics. 12+ computer centers. 
American Journal of Computational Linguistics Microfiche 43 : 5 
Behavioral, social sciences and related mental health fields 
Telephone calls, letters, and site visits. 
Prior to submission. 
Send draft of ideas, objectives, budget to 
Richard Lopez 
Center for Minority Group Mental Health Programs 
National Institute of Mental Health, Room 7-103 
5600 Fishers Lane 
Rockville, Maryland 20852 
Telephone: 301-443-3724 
American Journal of Computational Linguistics ~icrofiche 43 : 6 
The U.S. Government holds currency in these countries that 
cannot be exported. 
Joint programs in scientific and engineering research are sought. 
Support is available for 
International travel for project development 
and meetings 
Visits by individual U.S. scientists to 
institutions in participating countries 
American Journal of Computational Linguistics Microfiche 43 : 7 
CATASTROPHE THEORY 
Theme : Computation, Energy, and Ecology 
Address: H. B. Hair 
SIAM Headquarters 
33 South 17th Street 
Philadelphia, Pennsylvania 19103 
American Journal of Computational Linguistics Microfiche 43.: 8 
Cross-disciplinary theory and research in artificial intelli- 
gence, ccj,~itive and developmental psychology, individual 
differences, measurement, and instructional design. 
DIRECTORS Professor Joseph M, Scandura 
Merge Research Institute 
1249 Greentree Lane 
Narberth, Pennsylvania 19072 
DATES I To be arranged 
LOCATION r Banff Center, Banff , Albetta (probable) 
American Journal of Computational Linguistics Microfiche 43 : 9 
Humv aspects including hardware and software interface design, 
programming, interaction with different classes of user, 
training and modelling. 
DIRECTOR: Professor B. Shackel 
Department of man Sciences 
University of Technology 
Loughborough, Leicestershire, U.K. 
DATES r August 1976 
LOCATIONS Greece or Italy 
American Journal of Computational Linguistics Microfiche 43 : 10 
Alternate approaches and uses of computer-based instructional 
systems in University Science teaching: Games; models and 
simulation; computerized individualized aptitude evaluation 
and strategy development; interactive computer graphics; com- 
puter-managed instruction; multi-media instructional programs 
DIRECTOR: Professor A. Jones 
Universite Catholique de Louvain 
Celestijnenlaan 200-C 
B-3030 Heverlee. BELGIUM 
DATES : 19-30 July 1976 
LOCATION: Louvain-la-Neuve, BECGIUM 
American Journal of Computational Linguistics Microfiche 43 : 11 
CONGRESS 
AMSTERDAM - GRAND HOTEL KRASNAPOLSKY 
PRESIDENT: 
SPONSORS 1 
ADDRESS : 
Max H. Fisch 
Charles S. Peirce Society and Foundation 
International Association for Semiotic Studies 
Iilternational Union of History and 
Philosophy of Science 
Semiotic Society of America 
Institute for Studies in Pragmaticism, 
Texas Tech University 
Peirce Edition Project, Indiana University-- 
Purdue University at Indianapolis 
Research Center for Language and Semiotic 
Studies, Indiana University 
Carolyn Eisele, Chairman 
Douglas Greenlee 
Charles S. Hardwick 
Peter Hare 
Kenneth Ketner, Secretary 
Edward C. Moore 
Vincent Potter 
Nicholas Rescher 
Gresham Riley 
Thomas A. Sebeok 
Kenneth Laine Ketner, Secretary 
Inhtitute for Studies in Pragmaticism 
Texas Tech University 
P. 0. Box 4530 
Lubbock, Texas 79409 
806-742-1193, 2272 
American Journal of Computational Linguistics Microfiche 43 : 12 
NEW JOURNAL 
COGNITIVE SCIENCE 
EDITORS: Eugene Charniak, Allan Collins, Roger C. Schank 
CONTENTS 11lr PROBLEM SOLVING IN SEMANTICALLY RICH DOMAINS 
AN EXAMPLE FROM ENGINEERING THERMODYNAMICS 
R. Bhaskar and H. A. Simon, Carnegie-Mellon 
HUMAN AND COMPUTATIONAL QUESTION-ANSWERING 
W. Lehnert, Yale University 
DEFINITE DESCRIPTIONS G SEMANTIC MEMORY 
A. Ortony and R. Anderson, U. of Illinois 
ARTIFICIAL INTELLIGENCE, LANGUAGE. AND THE 
STUDY OF KNOWLEDGE 
I. Goldstein and S. Papert, MIT 
 EDITORIAL^ WHY COGNITIVE SCIENCE 
Allan Collins, Bolt Beranek and Newman 
"The discipline might have been called applied 
epistemology or intelligence theory, but 
someone on high declared it would be cognitive 
science and so it shall." 
American Journal of Computational Linguistics Micfofiche 43 : 13 
A new, standardizedthreeltep procedure has been announced by 
NSF for reconsideration of proposals initially declined. 
1. Program Director is to supply explanation on request 
2. Investigator is to request reconsideration within 
180 days after rejection of proposal 
3. Deputy Director is to review proposals on request 
within 180 days of reconsideration as in Step 2. 
Re-examkne administrative judgments 
as to scientific merit; can call for 
additional peer review 
No further review will be made following these three steps, 
but a new proposal can be submitted. "New" means "substantially 
revhed" 
Ametican Jburnal of Computational Linguistics Microfiche 43 : 14 
CONFERENCE CHRONICLE 
PAST AND FUTURE CONFERENCES NOTED FOR THE RECORD 
MIT - ATbT CONVOCATION: SESSIONS ON NEW APPROACHES TO A 
REALISTIC MOOEL OF LANGUAGE George A. Miller and Morris 
Halle, convenors. Bresnan, Fujimura, Forster, Kaplan, Wanner, 
Garrett, Blumstein, Zurif, Jackendoff, Carey; Maratsos. 3/9/76 
1 R 1 I ANALYSIS OF WRITTEN AND SPOKEN LANGUAGE AND ITS 
APPLICATIONS Morris Salkoff, director. Kayser, Kittredge, 
Gross, Haton, Miclet. Paris, 4/12-16/1976. 
ASSOCIATION FOR EDUCATIONAL DATA SYSTEMS - 14TH INTERNATIONAL 
CONVENTION Don Bitzer, Plato chief; Herb Grosch of Computer- 
world; contributed papers. May 4-7, 1976, at Phoenix. 
COMPUTER LAW ASSOCIATION - GROSCH AS LUNCHEON SPEAKER. 
San Francisco, April 22-23, 1976. 
SECOND INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING 
San Francisco, October 13-15, 1976. IEEE Computer Soc & ACM 
ACM SPRING SEMINAR April 12-13, 1976, at Chicago. Alex Orden 
U Chicago: Societal impact of computing. Peter G. Lykos, IIT, 
Computers in education: Where are we and where are we going? 
American Journal of Computational Linguistics ~icrofiche 43 : 15 
The 1976 
linguistic Institute 
oP the Linguistic Society 
STATE UNIVERSITY OF NEW YORK 
Director, Carol F. Justus Asalstant Director, Rand Bishop 
Associate Director. Wimred P. Lshmann GUNY G~o~dinabr. Francine Frank 
Kenneth Abrams, SUNY-Buffalo 
Morgan Jones, SUNY-New Palh 
Frank Anshen, SUNYStony 
Dennis Preston, SUNY-Fredonla 
Harold B-I FiUNY-Qeneseo 
Dmald Reiff, University of Rochester 
Robert Crornack, SUNY-Cortland WMlam Rltchie, Syracuse University 
Margaret Honwr, SUNY-Oneonlm Llnda Waugh, Cornell University 
Paul Hopper, SUNY-Blnahamton 
FACULTY 
MERVYN ALLEYNE 
FRANK ANSHEN 
ALAN BELL 
RAND BISHOP 
MARINA BURT 
LYLE CAMPBELL 
HEIDI DULAY 
CHARLES FERGUSON 
FRANCINE FRANK 
VICTORIA FROMKIN 
HOWARD GILES 
TALMY G IVON 
JOSEPH GREENBERG 
PEG GRIFFIN 
SHIRLEY HEATH 
JOHN HELFELDT 
REG I NALD HENRY 
JOAN HOOPER 
PAUL HOPPER 
CAROL JUSTUS 
EDWARD KEENAN 
STEPHEN KRASHEN 
SUSUMU KUNO 
WALLACE LAMBERT 
WINFRED LE HMANN 
CHARLES LI 
RICHARD LODER 
DENNIS PRESTON 
WILLIAM RITCHIE 
WILGA RIVERS 
DAVID ROOD 
ROGER SHUY 
SUSAN STEELE 
STEPHEN S rRAIGHT 
M, STUDDERT-KENNEDY 
DAVID TAYLOR 
SANDRA THOMPSON 
STANLEY WANAT 
LINDA WAUGH 
JOSEPI WIECHA 
MARIANNE WILLIAMS 
WOLFGANG WOLCK 
WALT WOLFRAM 
LADISLAV ZGUSTA 
LECTURES AND CONFERENCES 
LANGUAGE IN AMERICA 
Weekly; sponsored by National Endowment for the Humahitjes 
DISTINGUISHED VISITING SCHOLARS 
Five; sponsored by American Council of Learned Societies 
TYPOLOGY AND UNIVERSALS 
Both weekly; conducted by Lehmann and Greenberg 
SOCIETY FOR ELVISH STUDIES. JULY 10 
Lise Menn, 45 Dexter Road, Newtonville, Mass 02160 
STANFORD PHONOLOGY ARCHIVE WORKSHOP JULY 13-15 
Charles Ferguson, Linguistics, Stanford U., Calif 04305 
SECOND LANGUAGE ACQUISITION CONFERENCE JULY 17-18 
William Ritchie, Languages, Syracuse U., New York 13210 
HISPANIC AND LUSO-BRAZILIAN LINGUISTICS COLLOQUIUM JULY 24-25 
Joan Hooper, Linguistics, SUNY BufEalo, New York 14261 
INDO-EUROPEAN AND TYPOLOGICAL STUDIES JULY a8 
Paul Hopper, Linguistics, SUNY Binghamton, New York 13901 
AMERICAN DIALECT SOCIETY MEETING JULY 29 
Hood Roberts, CAL, 1611 North Kent St., Arlington VA 22209 
LSA SUMMER MEETING JULY 30-AuG 1 
1611 North Kent Street, Arlington, Virginia 22209 
TYPOLOGY AND SYNTACTIC FIELD WORK CONFERENCE AUGUST 2-4 
Stephen Anderson, Linguistics, UCLA, California 90024 
AMERICAN INDIAN LINGUISTICS CONFERENCE AND FESTIVAL AUG 6-7 
Lyle Campbell, Anthropology, SUNY Albany, New York 12222 
The 45th Summer Llngulstlc InstlWte ol the Linguistic Swiety of 
America is being hosted by the State University of New Yoh at Its Oswego 
campus. The institute has traditionally offered courses In particular 
aspeats of linguislics which are not readily available to students and 
faculty on their home campuses. 
The institute focusing on two major themes. "Language Varlatlon in 
America." In keeping with the bicentennial year, and "Languepe Univer- 
sals and Typology." will include oourses, lectures and seminars on these 
themes by internationally known scholsrs 
The Institute curriculum has been designed specificaliy to emphasize 
important Current developments In the dlaoipllne, the mutual contrlbutlons 
of linguistics and related disclpiines to one another, and the corn 
munication of these oontributions to linguists and noniinguists alike. 
Courses will meet Monday, Tuasday. Thursday and Friday for one hour 
and 15 minutes unless otherwise spedilied. An additionat 1 and tYI hours 
wr week is to be arranged in consultation with the instructor. Please con- 
sult the Institute flnal brochure tor course descriptions and prereduisltes 
Students who reglster for a course numbered 500 or above must be 
graduate students and prepared to do independent research in addition to 
following regular classroom work. Seniors may by petition to the 
Graduate Dean, be admltted to courses at the 5W level if they fulfiiCthe 
prerequisites, and show that they are capable of work at the graduate 
level 
COURSE OFFERINGS 
Courses include undergraduate and graduate offerings. In order to most 
effectively offer oourses for linguists, and for teachers and scholars out- 
side the field of iingui$tics, the Institute will offer ?-weak. Meek, Rwwk 
and bweek courses. 
The divlsion of the 8-week period 
into units of 2 weeks and 6 weeks 
is designed to glve linguists and 
nonlinguists an opportunity to ob- 
tain a structured sequence ot 
courses outside their particular 
area of specialization. A 2-week 
intensive Introductory course will 
be followed by 6-w& courses in 
areas ot interest to both linguists 
and nonlinguists. At the same 
time that students from other 
fields pursue the 2-week Intanstve 
introduction to Linguistics. 
linguists can aiu, take advantage 
01 courses in interdisciplinary 
topics Basic and advanced 
linguistics courses will be oflered 
as 4-week and Bweek courses. 
Courses listing only language 
requirements or introductory 
courses as prerequisites. and 
those without prerequisites, will 
be open to advanced un- 
dergraduate as well as to 
graduate students. 
Full-time student participants will register for 8 or 9 credlts Some typlcai 
combinations might include: 
July 9 
July 12 
.' 
(1) two &week coursqs 
(2) one 2-week intensive and two %we& courses 
(3) one 2-wdek intensive course, one 6-week and one &week course 
i4) one +weak intensive course. one 4-week course, and one bweek 
course 
No one may, however, register tor more than one Intensive course nor 
ontakr more than 9 semester hours of credlt Language arts teachers acd 
protaasional people In other fields am encwraged to take advantage of 
the 2-week plus Gweek structured sequence. Unguists are also en- 
couraged to explore related fields while pursuing more specialized in- 
terests in Rngulstlcs. 
JUNE 
July 23 
July 26 
AUGUST 
28 
20 
American Journal of Computational Linguistics Microfiche 43 : 19 
BAAL SEMINAR 014 
TRANSLATION 
DIRECTORS: R. R. K. HARTMANN AND W. 5. DODD 
SPEAKERS: K.-R. BAUSCH, BOCHUM 
i'. NEWMARK (CENTRAL LONDON  POLYTECHNIC^ 
K. REIS (WURZBURG) 
C. SMITH (CAMBRIDGE) 
AREAS: Linguistics-semantics and translation 
Training and craft of translator-interpreter 
Problems, procedures, aids to translatfion 
Translation in language teaching 
REGISTRATION$ Limited to 50 persons. Prefergnce to members 
of British Association for Applied Linguistics 
ADDRESS I The Language Centre 
University of Exeter 
Queen's Building 
The Queen's Drive 
Exeter EX4 4QH 
American Journal of Computational Linguistics Microfiche 43 : 20 
INTERNATIONAL SEMINAR 
INDEXING IN PERSPECTIVE 
WARSAW, POLAND 
21 - 25 JUNE 1976 
SPONSORz U N E S C 0 
ORGANIZER: National Federation of Abstracting & Indexing Services 
PURPOSES Training of working indexers from developing countries 
who wish to develop sophisticated information systems 
PROGRAM: History of indexing; systems and formats 
effects on the retrieval process 
LECTURERSr Everett H. Brenner, American Petroleum Institute 
Stella M. Keenan, U. of Technology, Loughborough 
Tefko Saracevic, Case Western Reserve University 
WORKSHOPS, Toni Carbo Bearman, NFAIS 
Gilbert R. Cintra, American Petroleum Institute 
ADDRESS: NFAIS 
3401 Market Street 
Philadelphia, Pennsylvania 19104 
American Journal of Computational Linguistics Microfiche 43 : 21 
OFFICERS : NATIONAL FEDERATION OF 
ABSTRACTING s INDEXING 
SERVICES - 1976-1977 
PRESIDENT: John E. Creps, Jr., Engineering Index, Inc. 
PRESIDENT-ELECT: Russell J. Rowlett, Jr., Chemica Abstracts 
DIRECTORSI H. E. Kennedy, BIOSIS 
Ben-Ami Lipetz, DocumentatLon Abstracts, Inc. 
Henry M. Koehler, American Dental Association 
Robert H.. Marks, American Institute of Physics 
SECRETARY, H. David Chafe, American Society for Metals 
TREASURER: Henry M. Koehler, American Dental Association 
RUSSELL J. ROWLETT, JR. is Editor of Chemical Abstracts Service. 
He was chairman of the Gordon Research Conference in 1974 and 
has been a member of the Committee on Chemical Information of 
the National Research Council for seven years. Before joining 
CAS in 1967, he was assistant director of the Virginia Institute 
for Scientific Research, and earlier director of research and 
development, Virginia-Carolina Chemical Corporation. 
American Journal of Computational Linguistics Microfiche 43 : 22 
MART1 N KAY 
Xerox Palo Alto Research Center 
3180 Porter Drive 
Palo Alto, California 94304 
SUMMARY 
The IBM 7040144 computer program described in this Memorandum 
is capable of applying a grammar consisting of unusually 
powerful rules to sentences in a natural language so as to 
discover their underlying structures All solutions allowed 
by the gramnar are found The Memorandum discusses the nota- 
tion used to write rules and the extent to which they can be 
made to state the same linguistic facts as a transformational 
grammar. 
NOTE 
This research was supported by the United States Air Force 
under Project RAND--Contract No. F44620-67-C-0045--monitored 
by the Directorate of Operational Requirements and Development 
Plans, Deputy Chief of Staff, Research and Development, Hq 
USAF. Views or conclusions contained in this Memorandum 
should mat be interpreted as representing the official opinion 
or policy of the United States Air Force. 
Originally published by The RAND Corporation, Santa Monica, 
California. as Memorandum RM-5452-PR, October 1967. 
Also published in the Proceedings of the International Confe- 
rence on Computational Linguistics, Grenoble, 1967. 
Any but the nost rudimentary tasks in automatic 
lanquage :>roce~sinq ncccssnrily involve o procedure which 
makes the grammatical structure of sentences c?xplicib and. 
cliscovcrs qrammatical a~1,iquitics. :.!ost exir;tinq programs 
are capable of applying context-free phrase-structure 
qrammars to the sentences anrl t!lc various techniciues for 
doinq this arc, by now, fairly well unrlcrstood. Ilowcvcr, 
this kind of grapnar is now qcnerally consitlered t6 be 
inadequate because it fails to revcnl t!le most important 
properties of the majority of sentences and becnuse it 
typically dctlarcs a sentence to be more amhiquous than 
it really is. These difficulties are overcome to some 
extent by context-sensitive grammars and to a large extent 
by transformational granmars. The proqram discussed in 
this :%emorandurn is capable of analyzinq sentences with 
context-sensitive qranmars and with grammars cf a class 
very similar to that of transformational qrammvs. 
Experimentation with tinis program is still in its very 
early stages, but the results reported here suqqest that 
the technique used may prove to he a powerful and efficient 
tool for language analysis. If so, it can hope to find 
application in informational retrieval, command and 
control. intelligence analysis and many other areas. 
AJCL will reprint from time to time contributions recognized 
to have signal importance in the history of computational 
linguistics and continuing value to students and practitioners. 
AJCL acknowledges with gratitude the approval of the copyright 
holders=-Martin Kay, The RAND Corporation, and Bernard Vauquois 
--end the extraordinary help rendered by RAND in ob~aining a 
photographable copy of this text.--Editor. 
TABLE OF CONTENTS 
(Introduction) 
1. The form of rules 
2. Phrase-s truoture grammar 
3. Transformational grammar 
Conclusions 
This ya?cr clcscribofi a soi7:>istic~tcA n\mr;iictic- 
andynis :lrogram for the IR 1 7040/44 con:>utcr and cliscusses 
unrestrictc?. rc7:~ritinq rulcs to a family of syr1:ol strinqs 
anA clcliv.~rs as out1:ut a11 t;le strinqn ti~nt can I)e RcrivcC 
fron !qcmi~crs of the initial family ;)y :~;c~ns.of the rulc~ 
provided. A sui~siciixy r.cch&nisq iloalo elit!l tho relation 
of dominance i.n the scnsc cn-:non in lin~uistics. This 
makes it possible for rulcs to rcfcr to co~:~lrltr? or partial 
syntnctic structures, or P-n*arl;ers, so that thc proqrnn 
can be used, at least to sone cxtcnt, for transCor~ationa1 
analysis. 
A program of this !:illr., \.~llic!i is intcnclcd for 
analyzing natural lanpngas, ~ust l~e ca?a'>le of opcrntinrr 
on a family of striagri as a ,s,inqle unit iiecaus", of tho 
gramatical anbiquity of :lards. Ta.l:o, for cxanyTe, t3e 
fanous scntence "Tirc flics like an arrow." Thcqe f!.vc 
words are not. t;lc:?selvos, thc nri~rary data on which n 
parsing proqrsn can ?,e ex.xxtcd to oL>eratc. Instcacl, 
each word is replaced 1,y onc ar norc oymlrols rcptcsentinq 
the grammatical catcgoricti to ~riiich it belongs. The 
assig~lments for t'nis exa:nrlc might !>o ::cn~:~r!i;lt as .fnllo:s: 
Word Grannatical catcqory 
Ti me 
flies 
like 
an 
arroT.l 
Noun, verb, asjective 
Plural noun, 3rd parson verb 
Singular noun, preposition, 
verb 
~ndcfihite article 
Singular noun, adjective. 
26 
TaXing onc catc<fory :;:f;d~ol for cnc:~ l.:ord, it is >osS lblc 
to form 36 i:ifTcrcnt strinc~s, ,srcsr-.rvil!g tbe ortler of the 
original scnt(?nce. Tllcse 36 :;txincja constitute t!lc family 
on :ihic!l the :>rogram ,ioultl o;)t?ratc if SC~ to nnnlyze this 
sentence. 
The program is said to perform as a non-deterministic 
clcvica !)ccouse, t*.r!~encvcr two rautually incom.:otil)lo rulcs 
are a,;~licrt:)lc to the nnric strin?, neither is nivcn Any 
:,ricrity; ?.>ot!~ arc ap!>licd, aiir; t:lo .rcsultinn ::trincrs 
ilcvalopcti inJc!?onAcntly. Civcn t:~c: ::t:r.i no "F T3 C" ~nc! 
tho rules 
the yrogran :.:ill thcrcfore .>roCuco nev strin~s: 
The program contains no mcc:lnnisi-. for c~unrdinq 
against seCucnces of rules which do not terninpte. If 
the granmar contains the follo:.rin? rulcs 
and the string to be parsed contains either "A B" or "B 
A", then the program will continue substituting these 
sub-strings for one another until the spce available for 
intermediate results is exhausted. This may cot seem to 
present any particularly severe proSle7 !~ccause a pair 
of rulcs suc:~ as CICSJ 1fru1A never ap car in any psogerly 
constructcc? qramar. gut, as :~c s:iall s:~ortly sce, 
cntircly ?lnusii)lc qramnxars can be constructed for vhich 
this !>roblcn rloaa wine. 
In order to c;ret a gcncrnl iilca of tl~c ~a:~ahilLtics 
of the program, it will LC uacful First to consir'.cr the 
 lota at ion UDCC! for ?rcsc?nting rules to it and thr, way this 
is interpreted by the nachixxe. In :!hot follo:vs, !re shall 
assume that thc roodcr is farnilinr :1it!1 thc tcrminololry 
and usual convci~tions of phrase-structure and 
transfornational grannar. An cxnm:)lc of the simplest 
kind of rewrit? rule is 
The "equalsn sign is used in place of the more familiar 
arrow to separate the left and riqht-l~nnd siclcs of the 
rule. The syml~ols on which the rules o~cratc are 'lords 
consisting of between one an6 si* c..lphaLctic c;lnroctcrs. 
Tho above rule will replace the ay-bol "WRST;" by a string 
of three synbols "P?r:S SC 1~:R.U" '.li~enevcr it occurs. The 
following rule will invert the order of the syd~ols "VEM" 
and "INC" 
VERB IMG = ING VERB 
The simplest way to represent a context free phrase 
structure rule is as in the following example: 
NP AUX VP = S 
?lotice that tile norm1 orAcr of the left and right-!land 
sides of the rule is reversed !zccause thc rccoqnition 
process consists in rewriting strings as single syn!>ols; 
the rules nust thcrofore tnlie t!le form of rccluctinns 
rat!~er than productions. 
The program vill nccopt p:?rnr;c structute rules in 
the form we have s:lown, but, 111 ai?:>lying t:~cn, it will 
not keep a record of the total eontcncr? structure to *.i:lic!i 
they contribute. In other words, it will caunc a nmi 
string to lje constructed, Sut will not rclnttl t'iix ?ti-ino 
in any way to thc string v'~ich %as rcwrittc'n. one l. c.. -.. 
to cause this rclatio~ship to lje :srcst?rvrd ic to .:rite 
the rule in the folloving forn: 
The numbers following the s:~::l,ols on the left-hand $id;-? 
of bhe rule function very much like the nunl,crs frcr:ucntly 
associated with structural inriices in txnnsforaatiqal 
rules. [?hen thz left-hand  sic;^ of t'lc rule il; fruvc! to 
match a particular sub-string, t!la nucljc- a%;<oci.~t.:r3 it'? 
a given s:rml~ol in the rulc :>ccoaes a ointcr to, nr a 
tem;>orary name for, t!~at slrn!-01. f:ith t'lis i:~t~-~!>.r,?t-~ti on, 
the left-hand sicle of t:~ z?ovc rulc can l.e read sc:-o':!~-lt 
as follows "Find an KP and call it 1; find an LUX follo:~ing 
this and call it 2; find'alVP following this and call it 
3." 
Thednud~ers in parentheses after a synbol on the 
right-hand side of a rule are pointers to itcns ilentifiz? 
by the left-hand side, and !!!lich the :lev sv~.\ol nust 
dominate. In the exam?le, thc sydjol "Sf' is to ::oril:nt@ 
all the symbols nentionecl on tha left-.lard siCe. 
A pointer nay refer to a r;inr;lo s:r?&ol, 8s we have 
Sho:m, or to a string of sy~mls. "!re Collowing rule is 
eciuivalent to the or.c just rlascribcc'. : 
Furthermore, the string to *rhich a .?ointer refers need 
not be continuous. Consi~?er tl:e Lollo~riny r?::nnple 
Thig will cause any string "!TP AUX VP" to be rcvritten 
as "S," hut the "S" :gill Qominate only "iiP" and 'I VP . " 
There be no evidence of the intcrvcninrj "AUX" in the 
final P-marker which wi11 contain the following ?!lrnse: 
Consider now tile follo~~inq ?air of rules: 
If these rulcs are a:splictl to the strinr~ "A R C D" the 
fo1lo:iing P-markcr will kc fornerl: 
Notice that the first rulc in the pair not only re-nyrlers 
the sy~bols in the P-marker but Forms two phr~sas 
~imultanoously. 
A different way of usinq L>ointer nur.!>ers on the 
right-hand side can be ill~~striltcr~ Sy co??nrinn the effects 
of the follorrinq trio rules: 
\\That is requireci, we assume, is a context scnsitivo ;>hrase- 
structure rule which will rewrite "P Sc" 2s ":,!OLCJ" in th8 
environment before "V SG". The first rule aci~ieves this 
effect but also inkroduces a new "V" dominatinq the old 
one, and a nev "SG". The secone rule ::oes what is really 
wanted: It constructs a phrase labeled "i\TOUiJ1' as re~uired, 
and leaves the sydols referred to by pointer number 2 
unchanged. 
The context-sensitive rule just considered Is 
presumably intended to insure that singular verhs have 
only singular subjects. A second rule in tlhich "SG" is 
replaced by "PL" would be required for i~lural verbs. 
But, since agreements of this kind nay  jell have to be 
specifled in other parts of the qrammar, the situation. 
might better be described by the fol1o:~ing three rules: 
The first two rules introtlucc J. node lnl~cled "llIJbI" into 
tile strnctwre nbovc? the sinqulrlr anc1 alural mnrp!len>cs. 
The third rulc cllccks for ~grecnent anit forms the subject 
noun phrase. Pointer ~?urnl~)or 2 is associated wit11 the 
symbol "NUPI" in the second place on the left-hand sic'e, 
and occurs :~y itself in the fourth ?lace. This means 
that the fourth symbol ~otchec! by the rulc ~.ust be 81~~~~88, 
and also that it must dominate c::actly the sane sub-tree 
as the second. In the exan2le we are assuminq that "P?L'il" 
governs a single noclc :::~ich vill :,e latclcd either "SG" 
or "PL" and tho rule will ensure that wilichover of these 
is dominated 5y the first occurrcncc of "PJTJ':" -:ill also 
be dominated by the second occurrence. Rotice that noun 
and verb phrases could be Eorxd ~iru1taneousl:r by tho 
folloriing rule: 
The synbols "ANY" and "IGULL" are treated in a special 
17ay by this progrhm and should not occur in strings to 
be analyzed. The use of the symbol "NULL" is illustrated 
in the rule: 
PPH = NULL 
This will cause the symbol "PPH" to be deleted from any 
string in which occurs. The program is nondeterministic 
in its treatment of rules of this kind, as elserrhere, so 
that it will consii.er nllnly:cs Z.n ~.r!iic:~ tllc svd>ol is 
Aelntcd, as r.leI.1 a5 any rrllicl~ can !;c isitdc :ay rateininq 
it. The syn:,ol "MULL1' is 11cc(: only r)n t'le right-'land 
sidcs of rules. 
r?!~e s:rn:>ol "ANY" is riscd nnlg an the lcft-;land sides 
of rules nnd has tile propi?rty ti~at t!lo vord in-li.as, 
namely thzt it will natch any s:r;?::ol in a strinr~. The 
use of this special synhol in illuntrntcd. in t!:e ?ollo;;ing 
rule : 
This will form a verb ~11rase frorl a Ver'~ clnd a noun ~hrase, 
with one intervening r~orcl or ?.lrzsc, rhnso rrrarrsaticnl 
category is irrelevant. 
Elements on thc left-hand sirTcs of rules can be 
specified as optional Sy writing a enllar sign to the 
left or right of the symi~ol as in tl~c folln1lj.n.; rule.;: 
The first of these forms a noun ;>;lrase from a deterhiner 
and a noun, with or vithout an intervening sdjcctive. 
The second is a new version of a rule alrcnrVy comsiciored. 
A verb p!lrase- is formed from a verb znd a noun phrase, 
with or. without an intervening :.rord or phrase of some 
othelr type. 
Elements can also be specified as repeatable by 
writing an asterisk against the synbol, as in the follo~ving 
example: 
This says that a verb ~ahrzse nay consist of a verb followed 
by one or more noun phrases. It is often convenient to 
be able to spccify that a given clenicnt rioy occur zero 
or nore tines. T!iis is Cone in the o!)vious ,.ray by combining 
the dollar sip and the asterisk as in the follorrinq rule: 
nccordinq to this, a noun may constitute a noun phrase 
by itself. llowever the noun may be prcccded b:r a determiner 
and any nurni3er of adjectivps, ancl follo~:recl. by a 
prepositional phrase, ancl all of t'lene :,?ill !~c e4,rocecl 
I>y the new noun phrasc Elat is formed. rrotice that the 
asterisk and the dollar sign can be placed 5efore or after- 
the symbol they refer to. The coml~ination is often useful 
with symbol '"NQY" in rules of tile followinq kind 
This ims similar to an earlier example. It cod>ines the 
number morpheme with a subject noun and vith a verb, 
provided that the two aqree, and allows for any number 
of other symbols to intcxvono. The symbol "LNY" with on 
asterisk and a dollar siqn corresponds in this syste? to 
the so called variabIes in the familiar notation OF 
transformational grammar. 
Consider now the foiloy~ing rule: 
This will form a noun phrase from a subordinating 
conjunction followed by a noun phrase, provided that this 
dominates only the symbol "Sn. Any symhol on the left- 
hand side of the rule may be followed by an expression 
in parentheses specifying the stxing of charaoters that 
this symbol must c:ircctly c?oninntc. This ex;,ression is 
constructed oxnctl;? 1i::c tile left-hand sir'cs cf rules. 
In particular, it may contain sy~:>ols Inllo..~er? by 
exizcssions in parenthosea. Tlle follo*.rinq rulc will serve 
as an illustration of this, nnr? of another nc:v Pcaturb: 
$ PP.5) = 1 3 4 TI14 DCF 4 D!: ADJ( (2) ) 5 
This rule calls tor a noun pt~rilsc consisting of a noun, 
a prcccding odjectivc wRich dominates a pre-ent partici?le 
and, optionally, a number of other clcncnts. This noun 
phrase is re~laccd by tile S,etcrminer from the Oriqinal 
noun phrase, if there is one, the clencnts i>rgceding the 
noun except for the prercnt ,nrtici-le, the noun itself, 
the synbol "!711", the zycol u~~:~T1, another  coy^ of the ' 
noun, the synbol "BE" , the ~y?i)ol "ADJ" clominntinq exactly 
those elexcnts originally clominated '>y "PPP??" an$, 
finally, any lollo~ring ?repositional p'lrases t!le oriqi~al 
noun phrase nay :lave contained. ?he nup!,cr "2" in double 
parentheses follo;.ring "I\DJ1' on tile right-!land side of 
this rule specifics that this sy~bol is to dozinzte, not 
the present participle itself, but t:~ clcncnts, if any, 
that it dominates. This device turns out to have ui2e 
utility. 
DouSle parentlleses can also be used folloving a 
symbol on the left-hand side of a rule, but with a different 
interpretation. Ne have seen how single parcntheses are 
used to specify the string immediately dominated by a 
given symbol. Double parentheses enclose a string vlhich 
must be a proper analysis of the, sub-tree dominated by 
the given symbol. A string is said to be a proper analysis 
of a sub-tree if each terminal syd~ol of the sub-tree is 
dominated by some member of the string. As usual, a 
symbol is taken to dominate itself. ns an example Of 
thin, consider the follovin~ rule: 
ART.1 S ( (>RT 17.2 ANY*) ) -1 2 = DET(1) 2 
This rule applies to a string consisting of an article, 
a sentonce, an2 a noun. The sentence qust bc analysal~le, 
at some level, as an article follower: hy a noun, follo~led 
by at least one other word or phrase. The noun in the 
eml>odded sentence, and the sub-tree it clorninntas , nust 
be exactly matched by the noun correspondinn to the last 
element on the left-hand side of the rule. The initial 
article and the em1xXidea sentence 7.iill 1x2 collecteri as a 
phrase under the sylnhol "DP:Tt' and the final noun wlll hc 
left unchanged. 
The principal facilities available for writing rules 
have now been exemplified. Another kind of rule is also 
available ~rhich has a left-hand side like those already 
des&ibeA but no "equals" sign and no riqht-hand side. 
Uowever it tiill be in the best interests of clarity to 
defer ap explnlmtion of !low these rules are interpreted. 
The user of the program may 1.1rite rules in exactly 
the form we have described or may add information to 
control the order in ~ihich the rules are ap;~licd. This 
additional information takes the form of an expression 
written before the rule and separated from if by a coma. 
This expression, in its turn, takes one of the following 
f orns : 
n1 in an integer which orders t!lis rulo relative to the 
others. Sincc tl~e sane ictcqcr can Sc nssi7llcd to nore 
than one rule, the ordering is ~artial. 9wlcs to :~:hich 
no number is explicitly assignecl arc qivcn t!lc nuni:cr 0 
by the program. n2 and ng, when l~roscnt, arc intcrpretcd 
as follo\;s: Evcry sy;-bol in the su5-stking m.atched by 
the left-hand side of t!le rule nust have !,em ;>reduced 
by a rule with nuder i, whcre n2~iln3. For these ?ur;)oses 
the sy~i:>ols in the orioinal family of strings offered for 
analysis are treated as thoug!~ they hac? been :?roAuced 11v 
a rule vith numi~er 0. 
2. PHRASE-STXUCTUPJ;: GPA!U.'A3 
It will be clear from what has been said already 
that this program is an exceedingly powerful device capable 
of operating on strings and trees in a vide variety of 
ways. It would clearly be entirely edequate for analyzing 
sentences with a context-free phrasc-structure granmar. 
But this problem has been solvec? !>cfore, and much more 
simply. We have seen how the notation can be used to 
write context-sensitive rules, and we should therefore 
expect the program to be able to analyze sentences with 
a context-sensitive grammar. However in the design of 
parsing algorithms, as else;ihere, context-sensitive 
grammars turn out to be surprisingly r?.ore complicated 
than contr?xt-free grannars. 
The ;>ro!~lcn t:mt context-sensitive r:rarq.mars pose 
for this yrogran can i~e s:~oi:n  i it!^ a ni?:>le exannle. 
1 
Consi2er tho follovring in 5ril:xnar: 
ABC (1) 
S 
I3 E (S) (2) 
This grammar, though trivial, is \!ell behave2 in all 
important aays. The lanquage rjcnerated, though reqular 
and unanbiguous, is infinite. 
Furthernore, every rule is useful for some derivation. 
Since the language generated is unaml~ir~uous, the gramnar 
is necessarily cycle-free, in other *:~ords, it ?reduces 
no derivation in which the sage line occurs more than 
once. Suppose, however, that the grammar is used for 
analysis and is presented with the string "A D E" -- not 
a sentence of the lanquage. Tho attempt to analyze this 
string using rules of tne gramnar results in a rewriting 
operation that begins as follows and continues indefinitely: 
'1 am indebted for this example, as for other ideas 
too numerous to document iridiviclually, to Susumu Kuno of 
IIarvard University. 
ADC 
A D I: (I>:? rule 6) 
utc. 
It would clearly I>@ possible, in principnl, to eciuip 
the program with a procedure for detecting cycles of this 
sort, but the timc? rdquirad by such a ?rocedure, and the 
compl'exity that it viould introduce into t!le nroqram as a 
whole, are sufficient to rule it out of all practical 
consideration. It might be argued that the strings which 
have to be analyzed in practical situations come from 
real texts and can he assumed to be sentences. The problem 
of distinguishing sentences from nonsentenccs is of 
academic ihterest. Rut, in natural lanquaqcs, the 
assignment of words to gramatical catc~ories is, as we 
have already remarked, notoriously a~?l~iquous and for this 
problem to arise it is enouqh for suitably ambiguous words 
to come together in the sentence. A sentence which t?ould 
be accepted by the above grammar, but which would also 
give rise to cycles in the analysis, might consist of 
words with the fol1o:iing graraaatical categories: 
The progran, as it stanr:~, cnntni~s :io ncc:l;lnisn 
which autonntically guards aqcinnt cyclcr,. I:o::ev~-?r, if 
the user !<noris tlhare thcy are li!:ely to occur or discovers 
then as a result of his ex;>ericnce ?ri.th t!lc ,?roqrnrn, !le 
can include sone s;>ecial rules in his qramar :!i~ici~ will 
prevent them from occurring. These rules, v;lich 1.r~ !lave 
already eluded to, are formally similar to all others 
except that they contain no "equals" si~n and no rig!it- 
hand side. 1*7hen a P-marker is found to contain a string 
which matches the left-hand sick of one of these rules, 
the program arranrjes that, thence fomr~~ard, no other rule 
shall be allowed to apply to the [thole string. The cycle 
in this latest example could not occur if the rjramlar 
contained the rule: 
ABE 
We now cone to the main concern of this paper which 
is to discuss the extent to which the program we have 
been describing can be made to function as a 
traflsformational analyzer. The main purpose of the 
examples that have been given is to show thd great power 
of the program as a processor of symbol strings. The 
notion of dominance is provided for, but only in a 
rudinentary :iay. It ccrtainlxr coulcl not bc claimed that 
the grogram is a trcc :)rocensor in any really 11or1:ahle 
sense. Rut gra~matical trannfornntions arc o:>crntions 
on trees and our invbfitinqtion t!larcforc: n:ust ta1:e the 
forh of siloy:ring that ti~cse ogcratinns can frequently, if 
not al:dys, :3e ni~ic1:ed :,y string rmvriting rules. 
IVe shall take it that a transPor11ationa1 grarnar 
consists of a context-Erco or context-nan:;itivc ?:lrasr?- 
structure conponent and a set of transformations orderec! 
in some way. To begin !jith, very little will ?-.e lost if 
we assume that the triinsformational rules .arc ni:~>ly 
ordered. 
Consider now the first transforriation in the list. 
In general, this may he expected to introduce 1:hrases 
into thc P-markers to which it applies vhich could not 
have been generated by t!le ;>hrasc-s tructure con~onent . 
Let us now write some additional phrase-structure rules 
capable of generating t2les.e new phrases. Let us izscrt 
these rules into the gramar in~eSiataly follovinq the 
first transformational rule 2nd establish the convention 
that, when they are used in the analysis ef the strinq, 
their output will be used only as input to the first 
transformation. Mow treat the second transfor~~ational 
rule in the sane way. It also can I>e cxpcctc~P to crcatc 
new kinds of phrase and phrase-structure rules can be 
written which woulZ recognize these. It nay be that some 
of the phrases formed by the second rule could also be 
formed by the first, and in this case, it may be possible 
to move.the appropriate rule from its position after the 
first transformation to a position after thc second and 
to mark it as ~roviding input only for these t~lo rules. 
Notice that the rules we are proposing to construct 
will not constitute what has sometimes been called a 
surface grammar. The phrases they rkscribe certainly do 
not belong to the base ntructura and ma!ly nF the- ma:* not 
be canable of surviving unri~anyrcc: into the surfacc 
ntructure. In general those rules r?csc?:il.,e ;'~rascs v:lich 
can only have transitory cxistcncc sor:c;t:-ir?re in thc 
generative groccss. ::oticc also t:-.nt in order to dcscribe 
these 9hrases adcc~uatcly it may somctiries !>e necessary 
to extend the notion of p1:rase structure grnnnar sometthat. 
C~nsicler for r?xan:?le the follo~rinq transformation: 
X-A-D-Y 
If ...re make the sual assunption that a rule is ap:>liorl 
repeatedly until no proper analyses of the P-marker rc~ain 
which can !>e matched by its structural index, tiler. this 
transformation, and many others, may ~>ro~luce phrases of 
indefinitely many types. ~et us su:,?ose that, before 
this transformation is applied for the first tine, all 
possible phrases that can he dominated hy the syn~bol "B" 
are describable by context free phrase structurQ rules 
of the follotiing form: 
:There the ai arc' ally strinqs. The ;)hranc structure grammar 
neeAed to ckscribe all the p:lrancs thzt can cxist after 
the operation of this trnnsfornation !.lust contain thc 
follominq ruler,, or narc n~onr~atol:. ruln schclv.ta 
where the asterisk indicates one or pore repetitions of 
the symbol "A". If the left and right-hand sides of these 
rules are reversed and tiley are ?resented to t!le proqrnn 
in the proper notation, t!len t!le transformation itself 
can bc rc2rescnted Ly the foiloving -air of rules: 
Since there are no facilit?es for specifying dominance 
relations among elements on the right-hand sides of these 
rules, it is necessary to resort to su1,terfuge. The 
phrase dominated by the synbol "B" is reproduced in the 
output of this rule with copies of the symbol "A" removed 
from the right-hand end and the remainder bounded by the 
symbols "B+" and "+B". These symbols serve to delinit a 
part of the string which can only figure in the complete 
analysis of the sentence if it constitutes a phrase of 
typo "But. The second rule ro.novcs these lmundnry sy??hols 
from the phrase of type "R" and, since no !~ointer is 
assirjncd to them, they will lsAve LO tracc in the final 
P-marker. 
Another, and perha2s mre economical, way to tlrkt'e 
recopnition rul-es corresponding to this transformation 
involves conflating the additional phrasc-strt~cture rules 
withthe reverse of the transformational rulo itself to 
give rules of the Ebllo~~ing kind: 
In fact, the elementary transformation for dau~hter 
adjundtion thatlwe are providing for here is more general 
than that often allowed by transformational grammarians. 
It is common ts require that if some element 5 is adjoined 
as a daughter of another element then 2 must hove no 
daughters before the transformation takes place. 
Sister adjunctidn can be treated in an analogous 
manner. Consider the following transformation: 
X-A-B-Y 
Adjoin 2 as right sister: of 4. 
The phrases existing before this transformation is carried 
out, and which have "Bn as a constituent, can be thought 
of as being described by a set of rules as follows: 
IIcre the ni arc nontcrr.inn1 syrl>ols and thc ai are strin~r,, 
possibly null. The qramar y9!lich :?cscri)>o:: the phrases 
existing after t!le o:)crntion of t!lis tral:!:For?-ntion nust 
contain, in aAdition, the following rules: 
The reverse transformation itself can now be represented 
by a set of rules as follows: 
Notice that the strings referred to !>y the synbols "X" 
and "Y" in both of the above transformations are unchanged 
by the transformation and are therefore not mentioned at 
abl in the analysis rules. Experience shm.is that it is 
in fact rarely necessary to ririte a separate rule for 
each ai. In nost cases, a transformation of this kind 
could he 1:ancllad in the progra.rn. .with a rule ot the follotlinq 
form:. 
Tilis is one of a largc 17nrher of cascs in vhich it has 
hcen found that the analysis rulcs can '~c made more 
permissive than the orininal qrammar sugqcsts ;iithout~ 
introducing s2urious structurcn and without serinusly 
incrensing the amount of tine or si?acc 11scc1 !'y the ?rocgram. 
vrhile it is possible that tr~nsfornationdl xnalynis 
can be done in an interesting :.ray :.qit!~ a procjram of t!lis 
sort, there soems to be little :lope of finclinq an alqoritlln? 
for writing analysis rules corresponding to a given 
transformational grammar. The followinq rule also involvos 
sister adjunction but ~oscs nuch more serious problems 
than the previous example: 
Adjoin 2 as right sister of 4 
The problem here is that a. variable "Y" intervenes between 
"A" and "B". On the face of it, the analysis rule 
corresponding to this transformation would have to be 
somewhat as follows: 
And in principal the program could carry out a rule of 
this kind. However the first symbol on the left-hand 
side of this rule will match any string \ihatsoever, so 
tl~ab, if the rule can bc applicc? at all, it can be applicd 
in a :?ro(iir~ious nuntxr of ::nys. nut, 1:ith rcal grammars, 
it usually turns out tllat quite a lot can he said about 
the part of the scntcnce covercd by the variahlc "Y" so 
that analysis rules can be vlrittcn which are snfficiently 
specific to be ~rncticablc. 
Deletions are notoriously troublcsoze in Trammars 
of any kind because t!lo:r can so onsily sive rise to cycles 
and undecidable prol>ler?s. Transfornational nrnmmarinns 
require that lcxical itcns silould only bc c?clateC from o 
P,-marker if thcrc is some other copy of thc snnc ite~ 
which remnins. This conclitioi: insurcs v:llnt t:~cy call the 
recoverability of thc transformation. ITw~cvcr, it is 
very important to realize that recoverability, in this 
sense, is a very *.ieak condition. The requirement is that, 
knowing that an itcn has been 6elctcd from a certain 
position in the P-mar::er, it shculd Se possil>le to tell 
what that item :;as. nut there is no rcc:uircnent that a 
P-marker should contain evidence that it !,as r?criveA by 
means of a deletion transformation or of the places in 
it where deletions night have taken ~lacc. 
Deletions are nore easy to cope vith in certain 
qituations than others. Considcr for exarlplc t!le followinq 
transformation: 
X-A-B-A-Y 
Delete 4. 
The recoverability requirement is satisfied because of 
the identity of the second and fourth elements in the 
structural index. The corresponding rule for the program 
It is liccessary to ;3rovirio ordcrinrr j.r.for?:ation vfith a 
rnlc of this kind !;ecousc it ::nuld otllervisc Ije capal>lc 
of oporating on its o!.rn output and c;rclin~ indefinitel:?. 
cut presuma!>ly this trans:ornation can !~o carried out any 
number of timcs'and the r,mc tll2rcforc c,:tould bc true of 
the corrcsponding analysis rule. Once again, experience 
shows that the grammarian alnost invarial~ly knows vorc 
about the cnviron~??cnt in which a delction talrcs place 
than is stated in the rulc, and if t!lis infor~nation is 
used carefully, arlalysis rul.es can :,e -,written :fhich do 
not lead to cycles. 
In principle the situation is oven :..orsc in rulcs 
of the following kind: 
X-A-Y-A-Z 
Delete 4 
Here the third elencnt is a variable ;i!lich con cover any 
number of nodes in the P maw. In analysis we are 
therefore not only without information about how many. 
times the rule may have been applied but we know nothing 
about where to insert new copies of the symbol "A", except 
that they must be to the right of the existing copy. 
The other commonly used elementary transformations 
(substitutions and Chomsky-adjunction) do not-present 
special problems. The main outstanding difficulty comes 
from the fact that transformational rules are ordered. 
We !~nve alrcacly said that the thecry of transformational 
qrammar is in a state of continual c!iacqe anc? this is 
particularly true of the ,>art that concorns the ordering 
of rules. For this reason :.!e IIRVC PSOURO? that the rules 
arc aiaply ordorcd in the hol~c tiint othcr :~ossi!~i~ities 
will not 1)c notaSly :‘lore cl.iEficult to 1:lcal with. Me shall 
also make the asnum:'ltion tknt transfornational rulcs are 
all obligatory. 
Consider npw tho follo~h~i~~cj qrar,nar 
Phrase st1:ucturc 
1. S -r A (D) B C 
2. C-DE 
Transformations 
and suppose that the program is required to analyze the 
string "A D B En. Since, in generation, the list of 
transformations is read from top to bottom it is reasonable 
to suppose that in analysis it should be read from bottom 
to top. v?e may take it that the analysis rule corresponding 
to the sccone tkknsformotion is somcl~hat ns follovs! 
"llis, togetllcr with the ti70 phrase-structure rules, is 
sufficient to give a comi~lctk analysis of tllr: strinq vith 
this underlying P-mar1:er : 
But if this is an unfierlying P-narkcr, the second 
transformational rule could not possibly be used to 7roduce 
a derived structure from it because the first 
transformation, which according to our assun!>tion is 
obligatory, can be applied to it qivinq the follo~iinrl 
result: 
It is in fact not sufficient to scan the list of 
transfornations from bottom to top because this procedure 
does not make allowance for the fact that the 
transformations are obligatory. To regard transformations 
as optional which were intended to be obligatory is in 
general to associate spurious base structures to some 
sentences. The solution for t?~e pre!;cnt cjranmar is to 
use the follo~iing set of a11a1ysis rules: 
Tho first ancl t!lircl rules contain, in effect, the str~~ctaral 
indices from tho second a112 first trnnuforxations 
respectively. The first rulc says that no strinq is 
acceptable as a sentence ~;1ic:1 contains "B 3" as a sub- 
string :~ecause to, this it ~iould have l~een :?ossihle to 
apply transformation 2. The sccond rule revcrsos t!le 
effect of transformation 2. The thir2 rulc ~>~lurle? any 
P-mar1:er existin9 at this stage riitll a ?ro?er analysis 
containing "A 9" as a sub-strina. T!lis is the structural 
index of transformation 1 which thcrcforc should have 
been applied to any P-~arl:er containing it. Vlc fourt:~ 
rule reverses the.effect of trnnsforr,ation 1 and the 
remaininq rules are the .hrnse-structure con:>oi;cnt of the 
granlmar. Once again it turns out tllat !,dlat nay 53. neccsaary 
in theory is only rarely ceccled in 2ractice. rx7ericnce 
with this progran is, so far, very linitc? !:ut no cases 
have so far been found in which incorrect analyses have 
resulted from omitting rules such as those nun!2cred one 
and three above. 
CONCLUSIONS 
-- 
It requires skill to write rules for nnalyzinq 
natural sentences wit11 the ->rogram described in this 
napcr. A rroqrar: can only ?roperly :>e calla? q 
transformational parser if it can :*or!. .!Lrcctl:r ,lith the 
uncr'litcc! rulcs of t:lc transfornatlonnl qramnar. lR~t no 
algorithm is known, nor is it 1il:ely tbat one dl1 s!~ortly 
;m fpund, ',r:lich \:ill proc;ucc fron n transeormational 
grammar a set of corrcs:>ontlinc rulos of the !:in<, rcc:uired 
by this yrogran. It is not diEEicult to cnnstruct a 
transfornational cfranmar for *.r:lich rio oxactly corresponc~ing 
set of ann1:rsis rules can :lo written. Ilo~iev~r, other 
grogrann have l,ccn s~rittcn s~!lich, t:-~ough t!~cy are' still 
in c!any ways iaperfcct, can more rcosonnbly bn called 
transformational parsers. l'hat then arc the advantages 
of the present proqram? 
The EUrrent version of the 2rogram is vrittcn in 
ALSOL and with very little regard for efficiency. But 
the basic algorithm is inherently a very rjrt?at Acal more 
efficicnt than any of its competitors. The various 
interpretations of an am!>iguous sentence, or a sentence 
which seems likely to be aml>iguous in the early stayes 
of analysis, are all worked on sinultancously. At no 
stage can the program be said to he developinq one 
interpretation of a sentence rather than anothat. If two 
interpretations differ only in some small part of the P- 
marker, then only one complcte P-marker is stored with 
two versions of the ambiguous part. Plork done on the 
unambiguous portion is Aonc only once for both 
interpretations. 
The program, though undoubtedly very powerful, seems 
naive from the point of view of modern linguistic theory. 
The program embodies very little of what tie know or believe 
to be true about the structure of natural languages. It 
might well be said that a computer program for analyzing 
natural languages is only interesting to the extent that 
it makes a claim about the basic form of those languages. 
Out the Lxorjram :?oscri!>ar: hero is intended as a to01 and 
not as a linnuistic hy:~othnsis. "!here is viuch to l>c 
learnec? a::out nzturnl lnnquaqe frbm ru~cinotinq on tho 
form of ullivorsnl nol:orntivc qrn?a..nr and tradiny countcr- 
axnnl?lc for cxainylc. 3ut thero is qlso rn~ch to be lcilrnad 
from studying tc:ct as it nctuclly occurs. Tho small 
amount of worl: that has so Far ')con ':one v.yit!l this procjran 
has hccn sufficient to auq~z??t stronnly thnt a sct of 
rulcs clerivec? algorit!l-7icnlly fron a tronsforr~ational 
qramar is unlikely to bo tho prost cffoctiva or the nost 
revealing analytic Zcvicc. 
American Journal of Computational Linguistics Microfiche 43 : 53 
d$p \(lashington Report 
AJCL EDITOR'S NOTE 
In August, 1975, the Washington Office of AFIPS began publishing 
a Report. By the time the Report's existence came to be known 
to AJCL, and arrangements for reprinting were settled, the 
first volume of the Report had been completed and the second 
well begun. 
We begin with the entire back file; henceforth the Report will 
appear regularly in AJCL, omitting only those items that AFIPS 
does not permit to be reprinted (for example, on account of 
restrictions imposed by third-party copyright owners). 
The news from Washington may be good or bad, but is' always 
important to those who live as close to the Federal Government 
as educators and researchers, not to mention engineers, must 
do. --DGH 
d! Washington Report 
................................... 
Amercan Feder:ltlon 01 Inf~rnlatl~lil Processing S(3csllus Illc. Sulle 4:'0 2100 L Sireel. N W Wdsli~nglon ll C 20037 
?O? (96 0590 
Vol. I, No. 1 August, 1975 
A MESSAGE FROM THE AFIPS PRESIDENT 
The first issue of this Washington Report signifies the real beginning of AFIPS 
activity in Washington. Through this medium we shall communicate to the AFIPS 
constituent societies and the members of these societies about computer and 
information processing related activities of the Federal government and about 
the matters to which the staff of the Washington Office is devoting particular 
attention. 
I am very pleased that the opening of the Wrtshington Office of AFIPS coincides 
so closely Mith the start of my own term as President of AFIPS. The establish- 
ment of this Office has been a particular interest of mine for some time; during 
my term as President I expect to devote a considerable amount of time to it. In 
order for the Washington Office to be a success, a number of pieces will have to 
fit together. Two of them most appropriate for mention here are: 
This Washington Report must develop into a useful, not to say 
indispensable publication to its recipients; this will happen 
most surely if we receive feedback from you on what you like 
and don't like, on what should be in the Report which is not 
and on what shouldn't be in the Report which is. 
The Washington Report is being distributed to senior officials 
of the AFIPS constituent societies. From you we solicit ideas 
on what the Washington Office should be involved in. Your 
suggestions on how the infonation processing comnunity can be 
best served by the AFIPS Washington Office are most important 
to us. 
Anthony Ral ston 
WASHINGTON DEVELOPMENTS 
DR. WILLIS H. WARE APPOINTED TO PRIVACY PROTECTION STUDY COMMISSION 
It was announced on June 10 that Dr. Willis H. Ware of the Rand Corporation has 
been appointed as a member of the Privacy Protection Study Comnission by the 
President of the United States. Dr. Ware will serve on the Comnission as an 
independent, private citizen. He is currently Chainan of the AFIPS Special 
Committee on the Right of Privacy, and was the first presitient of the Federation. 
In 1973, he served as Chainan of the Secretary's Advisory Conittee on Qutomat- 
ed Personal Data Systems, in the Office of the Secretary of the Department of 
Health, Education and Welfare; this committee produced the landmark report 
Records, Conlputers and the Rights of Citizens. 
Accordiog to the Privacy Act of 1974, the Commission iS to "(1) make a study 
of the data banks, automated data processing programs, and information systems 
of governmental, regional, and private organizations, in order to determine the 
standards and procedures in force for the protection of persolla1 ,information; 
and (2) recopv~end to tile President and the Congress the extent, if any, to which 
the requirements and principles of [the ~ct ] should be applied to the information 
practices of those organizations by legislaticpn, administrative action, or vol un- 
tary adoption of such requirements and principles, and report on such otlier legis- 
lative ~-ecoirtnendations as it may detennine to be necessary to protect the privacy 
of ilidividuals while meeting the legitimate needs of governnlent and society for 
infonnation.' 
NEW EXECUTIVE DIRECTOR OF DOMESTIC COUNCIL COt1MITTEE ON PRIVACY 
-..-- - 
Vice.President Rockefeller' announced on June 26 the appointment of Quincy Rodgers 
as Executive Director of the Domestic Council Committee on the Right of Privacy. 
I.lr. Rodgers is currently Minority Council to the Subco~luiiittee on Separation of 
Povrers of the Senate Judiciary Committee. 
Tile Ooiilestic Council Privacy Cornittee was established in 1974 as the arm of 
the Administration responsible for developing and coordinating agency views, 
pol icy recommcndations, and specific legislative and administrative initiatives 
regarding the collection, storage, and dissemination of infonnation about individ- 
ual hieri cans. 
EGER OLITLINES OTP AGENDA 
- 
John Eger, Acting Director of the Office of Telecommunications Policy, in the 
Executive Office of the President, outlined the agenda of his office in a 
I-lay 20th speech before the Computer and Business Equipnient Manufacturers Asso- 
ciation, Addressing what the Office of Teleconiii~uni cations Policy might do to 
insure a nlarket minimally constrained by governllient regulation, Egcr submitted 
the following: 
"One, encourage an even greater role for competition and market 
fzes in the com~nunications indbstry . . . . Two, address the 
privacy. and privacy-related problems intensifiny the sheer 
speed and eiffictency of computer coirmunications . . . And, 
three, . . . seek to maximize reliance on the private sector 
- 
for the provision of infonnation services . . . ." 
EFTS LEGISLATIVE PROPOSALS 
Against the background of the broad EFTS moratorium proposed last January by 
Senator Proxmire in the Senate and Mr. St. Germain in the House of Representatives, 
two bills of lesser scdpe have recently been introduced. The earlier Proxmire 
moratorium would require that no financial institution enter into a contract 
involving electronic metbods of fund transfers at places of business other than 
financial institutions and clearinghouses. 
Senator McIntyre introduced on June 6 a bill whYch would restrict certain 
Federal financi a1 institutions from engaging in transactions with its customers 
outside the state in which the Federal institution is located, unless the laws 
of that state specifically authorized such transfers. Mr. St. Genain intro- 
duced in the House on June 18 legislation which would impose a 90 day moratorium 
under which no Federal regulatory agency or financial institution could approve 
or authorize the establishment or expa%ion of any electronic funds transfer 
systenls; during this period, the National Com~ission on Electronic Furid Transfers 
would submit a report to tile Congress wiiicl~ would contain its r~conunendations 
regarding further legislation. 
EFTS APPOINTMENTS REMAIN PENDING 
Presidential appointments to the National Coniinissioii on Electronic Fund Transfers 
remain pending, although the Coiiriiission was created in October, 1974 bv Act of 
Congress. Concern over the delay has bee11 voiced by Senator FlcIntyre, Chairman 
of the Financial Institutions Subconr~~ittee of the Senate Banking Conmittee, as 
well as by Inembers of the private sector. The Commission is to recorntend, after 
a thorough study and investigation, appropriate administrative and legislative 
action in connectio~? wi ti1 the possible development of pub1 ic or private electronic 
fund transfer systems. 
PUBLICATION GUIDELINES FOR THE ;RIVACY ACT OF 1974 
The Federal Register on June 19 published guidelines for Federal Agency compli- 
ance with the Privacy Act of 1974. The OFR guidelines state that "[a] Federal 
Agency may not use a system of records after September 27th unless it has 
published in the FEDERAL REGISTER a notice for that system that meets the 
requirements of . . . the Act. Accordingly, the OFR guidelines require in each 
notice: the system name; location at which records are maintained; categories 
of individuals on irlion~ records are maintained; type of information contained in 
system records; routine use of the system; title and address of the agency 
official responsible for system; and a [means by which an individual can gain 
access to and contest records. 
FORD ADMINISTRATION INTRODUCES BILL FOR WHITE HOUSE SCIENCE ADVICE 
Senator Moss, in behalf of the Ford Administration, introduced on June 20 a 
bill to establish a science advisory mechanism in the White House. While there 
are currently several bills pending before the Congress on this subject, the 
administration bill (S.1987) and a bill introduced in the House by Mr. Teague 
(H.R.4461) are considered to be the most likely contenders for enactment. 
S.1987 would establish in the Executive Office of the President an Office of 
Science and Technology Policy. The Director of the office would be the 
President's chief policy advisor with respect to scientific and technological 
matters. 
THEODORE PUCKORIUS NPMED COMMISSIONER OF ADTS 
The Administrator of the General Services Administration announced on May 19 
that Theodore D. Puckorius has been named Commissioner of the Automated Data 
and Telecomunications Service (ADTS) ADTS is responsible for government- 
wide ADP resource management and procurement. 
AFIPS IN WASHINGTON 
AFIPS OPENS WASHINGTON OFFICE ON JUNE 26, 1975 
The opening of the AFIPS Was.hington Office, approved by the AFIPS Board of 
Directors in November, 1974, took place formally on June 26, 1975. 
Office Enarter. The charter of the Washington Office, as established by the 
AFIPS Board, is: 
(I) ,to provide an information service to AF!PS constituent 
societies ; and 
(2) to establish contact with Federal agencies and the Congress, 
and make experts and expertise available from the AFIPS 
constituency to such groups. 
Washington Activities Committee. The Washington Office will operate under the 
AFIPS Washington Activities Committee, chaired by Mr. Keith Uncapher, Director 
of the University of Southern California Information Sciences Institute. 
Committee members are Dr. Frank Ryan, Director of Information Systems for the 
U. S. House of Representatives, and Mr. Ralph Leathetman ol the Hughes Tool 
Company. 
Special Inaugural Presentation. The Washington Office was formally inaugurated 
by a conference addressing the theme of "Inforrsation Processing as a National 
Resource." Speakers, representing the Federal government, academia and 
industry, included Senator John Culver (Iowa); Dr. Brockway McMillan, Vice 
President, Military Systems, Bell Telephone Laboratories; Dr. Ruth Davis, 
Director of the Institute for Computer Sciences and Technology, National 
Bureau of Standards; and Dr. Allen Newell, University Professor at Carnegie- 
Mellon University. 
MARCENE TERRONES JOINS AFlYS STAFF 
Marcene E. Terrones has joined the P.FIPS staff as secretary in the newly opened 
AFIPS Washington Office. She brings matching credentials to a job which demands 
a bruad range of experience; most recently, she has served as Assistant Adminis- 
trator in the International Union of Operating Engineers and Pipe Line Employers 
Health and Welfare Fund. We trust you will all come to know her as you come in 
contact with the Washington Office; congratulations in your new job, Marcie, and 
we1 come aboard! 
PhlllpS Nyborg Director 
Washlngfon Offtce 
d% Washington Report 
************************************* 
American Federat~on of lnformatlon Processing Soclelles. Inc Sulle 420, 2100 L Streef. N W Washington D C 20037 202-296-0590 
Vol. I, No. 2 
WASHINGTON DEVELOPMENTS 
September, 1975 
NEW LEGISLATIVE ACTIVITY ON WHITE HOUSE SCIENCE ADVICE 
Representatives Teague and Mosher have introduced a new bill .(H.R. '9058)to pro- 
vide for scientific advice to the President, which integrates some major features 
of two earlier bills introduced by Teague and the Ford Administration, respective- 
ly (Washington Report, 8/75). 
Back round. There have been various science advisory mechanisms in the White 
+ ouse over the years, at times having a recognition of information processing. 
In 1957 President Eisenhower, in response to international developments. in mi s- 
sile technology, appointed a Special Assistant to the President for Science and 
Technology, and a President's Science Advisory Comi ttee (PSAC) which consisted 
of non-government scientists and engineers; in 1959, he also established the Fed- 
eral Council for Science and Technology, which was composed of representatives of 
the major science-oriented agencies. President Kennedy, in 1962, institutionalized 
White House involvement in science policy by creating the Office of Science and 
Technology (OST) in the Executive Qffice of the President. As of 1970 OST included 
a staff member for computers, Dr. A. Michael Noll. However, President Nixon abol- 
ished PSAC and OST in 1973, transferring OST's functions (except those relating to 
national security) to the Office of the Director of the National Science Foundation, 
Dr. H. Guyford Stever. While Stever was simultaneously appointed Science Advisor 
to the President, the post was organizationally removed to agency level, and Stever 
was effectively barred from White House policy level discussions during the Nixon 
Administration. The Ford Administratian, in contrast, has recenxly shown several 
signs of willingness to reinstate a science advisor in the White House. 
,H.R. 9058 differs from earlier bills in two important respects. First, in contrast 
to the Administration bill and to Stever's present role, H.R. 9058 would specifi- 
cally direct the new science advisor to render advice in the area of national secu- 
rity. Second, in lieu of the maior organizational reform proposed in the original 
Teague Bill (H.R. 4461) to provide a centralization of scientific research and 
development, the new bill would simply initiate a major study to assess the total 
Federal science and techology effort, devoting particular attention to organiza- 
tional reform. Since the earlier Teague Bill would have consolidated major Fed- 
eral scientific and technical information organizations, the study approach is 
expected to create considerably less political resistance. Nevertheless, the bill 's 
prospects for enactment remain unclear. Although AFIPS and other scientific soci- 
eties were told by an official of the Domestic Council in a White House briefing on 
July 30 that the bill was perceived to be a good one, it has not yet received the 
public support of the Ford Administration. 
PRIV4CY COMMISSION ELECTS CHAIRMAN, APPOINTS EXECUTIVE DIRECTOR 
The Privacy Protection Study Commission elected David F. Linowes as its chairman 
on July 11. Linowes subsequently announced on August 14 the appointment of Carole 
W. Parsons as executive director of the Commission. Mr. Linolves is an internation- 
al management consultant and a certified public accountant. Ms. Parsons has for- 
merly been associated with the White House DonieStic Council Committee on the Right 
of Privacy and the HEW Advisory Committee on Automated Personal Data Systems. 
In carrying out its statutory charter (Washington Re ort, 8 75), the Comlnission 
will emphasize the assessinent of privacy safeguarhe (usiness and industry 
sector. It will examine policy issues associated ~ith the interstate transfer of 
information, the use of personal identifiers, and information activities in areas 
such as employ~nent, insurance, credit, banking, health, and social service. It 
will publish a final report by June, 1977. 
Chairman Linowes has indicated "the strong desire of tlie Conaiission to receive 
suggestions from the public on which facets of the private sector should be given 
major attention;'' further, Ms. Parsons has stated that she will be looking for 
new ideas and lines of research to propose to the Conirnission. Tlie AFIPS Washington 
Office will endeavor to play an active role in providing assistance and input to 
the Commission; comments from tlie AFIPS societies are therefore of substantial im- 
portance, and are strongly solicited. 
U. 5. SUPREME COURT TO REVIEW NEW CASE ON PATENT OF SOFTWARE, 
There is now pending before the U. S. Supreme Court a review of the software pa- 
tent granted to Thomas R. Johnston. The Court ruled in the 1972 Benson-Tabbot 
case that the software patent then in question was not valid, but the Justices 
did not resolve the issue of software patentability in general. Instead, the 
Court deferred to the Congress for a definition of the legal protections to be 
afforded to software; the Congress, however, has not responded. In the brief for 
the United States filed in the pending case; Solicitor General Robert Bork has 
asked the Court to resolve whether a coniputer program is patentable as a machine 
(rather than as a process), arguing that it is unreasonable to characterize a 
programmed general purpose coniputer as a "new s~achine". 
MAJOR FEDERAL AGENCIES PROPOSE REGULATIONS T4 IMPLEMENT PRIVACY ACT OF 1974 
Federal agencies are required by the Privacy Act of 1974 to publish, prior to 
August 28, 1975, proposed implementing regulations for public comment. Comments 
from the AFIPS societies may be submitted through the AFIPS i.!ashington Office. 
Major agencies which have proposed rules subject to comment during September include: 
Deadline for Federal Rejz 
Agency Comnents Citation fpr Text of Rules 
National Science Foundation 9/5/75 7/29/75 p. 31811 
Securities Exchange Commission 9/12/95 
8/15/75 p. 34417 
Office of Management and Budget 9/12!75 8/14/75 p. 34165 
Health, Edvcation and Welfare 9/15/75 
8/14/75 D. 34129 
GENERAL ACtOUNTING OFFICE REPORTS WASTE IN COMPUTER SYSTEM ACQUISTIONS UNDER 
RDERAL GRANTS 
The General Accounting Office has submitted a report to the Congress which finds 
that, due to the lack of appropriate Federal procedures, recipients of Federal 
grants for acquisition of data processing equipinent have been allowed to utilize 
unsatisfactory procurenlent practices. The report stated that grantees: obtained 
new computer equipment without thoroughly evaluating their needs; obtained their 
own computer systems witho~t adequately exploring opportunities for joint use of 
existing computer facilities; leased equipment for short periods of time without 
fully considering the savings from purchasing or long term leasing; and consider- 
ed primarily mainframe manufacturers as sources of equipment supply, while giving 
little consideration to reduced-price sources such as leasing firms, used computer 
suppliers, and independent manufacturers of peripheral equipment. The report re- 
comends that Federal agencies establish consistent and economical acquisition pm- 
cedures, and that grantor agencies be required to ensure grantee cooipliance. New 
procedures may indirectly impact private (e.g., university) as well as state and 
1 ocal government grantees. 
NBS AND ADIS-TO JOINTLY DEVELOP COMPUTER INTERFACE STANDARDS 
The Institute for Computer Science and Technology (ICST) in the National Bureau 
of Standards (NBS), and the Automated Data and Telecammunications Service (ADTS) 
in the General Services Administration (GSA) have announced that they will joint- 
ly undertake an effort to develop computer interface standards; one desired re- 
sult of the program is to facilitate the competitive Federal procurement of com- 
puter peripheral equipment. 
ICST's overall charter is to develop standards related to information processing, 
and ADTS is responsible for automated data processing equipment procurement 
throughout the Federal government. The NBSIGSA program will: assess Federal 
government ,experience with multivendor computer systems and produce a report to 
the agencies; produce technical guide1 ines to assist Federal agencies in the plan- 
ning, selection and operation of multivendor systems; analyze comparativeLy the 
device level interface employed in high density disk drives; and evaluate the 
technological and economic impact of computer interface standards in the iiterna- 
tional , national and Federal sectors. 
NBS SOLICITS COMMENTS ON DATA,ENCRYPTION ALGORITHM 
Pursuant to its statutory authority to establish Federal ADP Standard?, the 
National Bureau of Standards is developing a standard data encryption algorithm 
to be recomnended for Federal use. The purpose of the algorithm is to allow for 
compatibility of encrypted data. NBS points out that cryptograghic devices im- 
plementing the standard may be covered by patellts held by IBM, which has agreed 
to grant non-exclusive, royal ty-free 1 icenses; and that devices re1 ating to the 
standard may be subject to ex~ort cont~l~ AFIPS societies may submit comments 
to the Associate Director for ADP Stagdards, Institute for Computer Sciences and 
Technology, National Bureau of Standards, Washington, D. C. 20234, or to the 
AFIPS Washington Office. Deadline fbr receipt of comments by NBS is October 30, 
1975. The specification for the data encryption standard may be found in the 
Federal Register of August 1, 1975 at p. 32395; alternatively, copies may be 
obtained through the AFIPS Washington Office. 
RECENT EFTS DEVELOPMENTS 
The bill introduced in the llouse by Rep. St. Germain (Washington Report, 8/75), 
which would limit installation of remote teller terminals to a financial insti- 
tution's ho~ne state and within 25 miles of its headquarters or 10 miles of a 
branch, has apparently died in conmiittee. While the measure to limit installa- 
tion of such ternlinals has been, therefore, essentially rejected by the Senate, 
a Federal district court in Washington ruled i~ July that remote teller terminals 
constitute branch banks and are subject to laws governing branch banking. The 
decision, if it withstands the appeal being requested in the U. S. Comptroller of 
the Currency, will effectively prohibit future ban< installation of such termi- 
nals, and will cause operation of existing installations to be discontinued. 
NEW BILL INTRODUCED TO CONTROL FEDERAL CRIMJ_NAL JUSTICE INJORMATION SYSTEMS 
- 
Sen. Tunney and Rep. Edwards have introduced identical bills (H.R. 8227 and 
S. 2008) to control Federal criminal justice information systelns. The bill is a 
compromise incorporating features of legislation introduced earlier this year. 
It would allow only conviction records to be distributed without restriction to 
law enforcement agencies; arrest and acquittal records could be disclosed only 
for specific purposes (e.g., the investigation of a specific crime] and only to 
authorized personnel. Further, the bill would create a Commission on Criminal 
Justice Information, composed of rnen~bers of state and local law enforcement 
agencies, which would implement and oversee enforcement of the act. I,lhile tile 
FBI has been attempting for some time to establish a law enforcement communicationr 
system interconnecting local police departments throughout the country, the Cormis- 
sion would have the authority to modify or terminate the FBI plan. 
AIR FORCE DESCRIBES EMPHASIS ON SOtTWARE ENGI'NEERING 
General! Sa~n Philliijs, Air Force Systems Comn~ander, bas described a new major em- 
phasis on software engineering in the procurement of weapotis-systems computer 
software by the Department of Defense. The approach described would require soft- 
ware to be engineered in niuch the same way that present procurements engineer 
hardware; greater effoFt would be devoted to work breakdown structures, ~nilestones, 
technical review of contract work, and software reliability assessment. 
FEDERAL GOVERNMENT INFORMATION SOURCES FOR CURRENT RESEARCH IN INFORMATION 
PROCESSING 
-- 
AFIPS societies mould be aware of the information resources of the Smithsonian 
Science Information Exchange (SSIE) and the National Technical Information Ser- 
vice (NTIS). The former organization provides information on ongoing or recently 
completed research projects, the latter on recently published research reports; 
both have developed on-line data bases. 
In the area of information processing, SSIE receives project des'c'riptions from the 
National Science Foundation, the National Aeronautics and Space Administration, 
the Department of Defense (unclassified data), and other public and private groups. 
The active file covers the 1 ast two government fiscal years, and contait.~ descrip- 
tions of approximately 1300 projects in computer hardware research, 000 in comput- 
er software, and 450 in information science theory and applications. In most cases 
the record for each project contains a 200 word description of'the work to be per- 
formed. Information services include custom searches performed on-line by the 
SSIE staff, ($50.00 for the search and first 50 records, $10.00150 records there- 
after) and Research Information Packages (pre-designed searches - about $35.00). 
Contact SSIE at Room 300, 1730 M Street, N.W., Washington, D. C. 20036, telephone 
(202) 381-4211, for further information. 
The NTIS on-line data base consists of Federally sponsored research reports conl- 
pleted from 1964 to date. Each record represents a Technical Report Summary aver- 
a ing 250 words. NTIS information products include the custom on-line NTISearch 
(!100.00 for the search and first 100 records, $25.00/100 records thereafter), and 
Published Searches (standard NTISBarch's - $25.00). NTIS also publishes Week1 
Government Abstracts, which carry summaries of research reports as they ade 
-1ic; subjects include: Computers, Control & Information Theory; 
Comnunication; and Library & Information Sciences. Contact NTIS at 5285 Port 
Royal Road, Springfield, VSrginia 22161, telephone (202) 967-4676, for further 
information. 
NEWS BRIEFS 
- After heavy debate on both NSF funded projects and the peer review system, 
Congress has passed the NSF authorization bill for FY76; the bill passed 
without the proposed Bauman amendment which would have required prior Con- 
gressional approval of NSF research grants. The FY76 authorization is a 
2.6% reduction (in absolute dollars) from FY75, but NSF transferred about 
6% of its research to ERDA before FY76 began. 
- OSIS and NTIS have published the Federal Scientific and Technical Comnu- 
nications Activities: 1974 Progress Report. Order No. PB-242 318, $5.25 
NTIS; Dept. of Commerce, Springfield, Virginia 22151. 
- Hearings. w. of Commerce: Domestic and International Business Adminis- 
tration, Computer Systems Technical Advisory Committee, Technology Transfer 
Subcommittee, September 9; Licensing Procedures Subcommittee, September 10; 
Hardware Subcomnitt'ee, Septemben 16. National Bureau of Standards: Federal 
Information Processino Standards Coordinatina and Advisorv Committee. Seo- 
tember 17. Cowtact A~PS Washington Office Tor details. - 
- Rep. Staggers has introduced legislation which would authorize new telecom- 
munications R & D. 
- Rep. Morris Udall has proposed legislation which would create a Com etition 
Review Conmission to assess competition in major economic sectorbg 
"electronic computing and communication equipment." 
AFIPS IN WASHINGTON 
AFIPS BECOMES CORRESPONDING SOCIETY OF NATIONAL RESEARCH COUNCIL 
AFIPS President Anthony Ralston has accepted an invitation from the National Re- 
search Council (NRC - the principal operating agency of the National Academy of 
Sciences and the National Academy of Engineering) to become a corresponding soci- 
ety of the newly-formed Assembly of Mathematical and Physical Sciences (AMPS).. 
AFIPS was formerly an NRC affiliated organization, but a recent NRC reorganiza- 
tion has required a restructuring of prior society relationships. Affiliation 
of AFIPS with one NRC assembly does not preclude affiliation with others, such 
as the Assembly of Behavioral and Social Sciences. 
In NRC tradition, AMPS activities will consist of both direct responses to re- 
search requests initiated by Federal agencies, atid the undertaking, on its own 
initiative,of projects which are deemed to be of national interest. In making 
the invitation to AFIPS, AMPS Chairman Norris E. Bradbury said,"it is essential 
that serious consideration be given [by AMPS] to both the benefits that can accrue 
to the Assembly Frogram through interaction with tlie professional organizations. .. 
[and] contributions that the Assembly, in turn, can ~nake to these organizations." 
AFlPS SOCIETY PRESIDENTS ASKED TO DESIGNATE WINGTON OFFICE LIAISON 
---- -...---- 
AFIPS has asked its society presidents to each designate a liaison to the AFIPS 
Washington Office. Persons designated will act as a primary working contact 
for the Office, providing access to subject area experts within each society, 
providing nominees for various pamls, relaying disseminated information to 
appropriate points within each society, and providing feedback regarding the 
Office and the issues with which it deals. 
ACIIPS WASHINGTON ACTIVITIES COMMITTEE MEETS TO CONSIDER WASHINGTON OFFICE 
~ETS 
In responding to its fundamental charter to facilitate conimunication between the 
AFIPS societies and the Federal government, the AFIPS Washington Office will be- 
come involved in certain project-oriented activities. The AFIPS Washington 
Activities Comnittee met in \dashington on July 23 to consider a list of 11 unso- 
licited proposals (including several on a funded basis) received from Federal 
agencies, for projects which the AFIPS Washington Office ]night undertake, initi- 
ate or facilitate. As a result of the meeting, 6 of the proposals are being 
furtliar investigated, for possible final approval by the Committee. In addition, 
presidents of the AFIPS societies have been asked to submit simil ar proposals. 
COMMENTS PLEASE! 
-- 
The content and format of the AFIPS Washington Report are in a formative and 
experimental stage. Your comments will be of great assistance in making this 
publication responsive to the needs of your society. 
AFTPS societies have permission to use material in th'e AFIPS Washington Report 
for their own publications, except that where an article title app ~s with 
an (*) c7earance must first be obtained from the AFIPS Washington 8fice. 
d! Washington Report 
**********i************************** 
American Federat~on of Information Proasslog Soctel~es Inc Sule 420 2100 L Street N W Wsshlnglon, DC 20037 202.296.0590 
Vol. I, No. 3 October, 1975 
WASHINGTON DEVELOPMENTS 
COMPUTER RESEARCH MAINTAINS STATUS IN NSF REORGANIZATION 
Computer research has maintained its status through the course of the recent major 
reorganization in the National Science Foundation (NSF). Under the new organiza- 
tional structure, computer research will be conducted IR the Computer Science 
Section (CSS) of the Directorate for Mathematical and Physical Sciences, and Engi- 
neering (MPE). The reorganization is designed to provide a more effective manage- 
ment structure, and to balance research activity among the new Directorates. 
Impact of the Reorganization. Dr. John R. Pasta, director for the former Division 
of Computer Research, will assume broader responsibility as the director of the 
new Division of Mathematlcal and Computer Sciences. Assuming Dr. Pasta's earlier 
responsibilities, Mr. Kent Curtis will become the new head of the Computer Sciences 
Section. According to Dr. Pasta, the recognition of computer science reflected In 
the reorganization will help to insure that it continues to be viewed as a unique 
discipline within NSF; the $12 million budget for research in this area will rain 
unaffected. 
Grants for computer research. NSF awards grants to support research in Computer 
Science, Computer Engineering, and Cmpl~ter Applications. CSS, In an effort to 
assess prior1 ties as seen from researchers in the field, will continue the policy 
of accepting unsolicited proposals from any source. While universities are expect- 
ed to be primary recipients of grants, other organizations are ellgible for support. 
Within CSS, priorities in research funding are reflected in the various programs, 
and program directors are the primary decision makers in determining grants; these 
individuals should be considered the primary contact within NSF for researchers 
interested in specific grant areas: 
- Theoretical Computer Sciences Program, Bruce H. Barnes 
- Software and Programing Systems Program, Thmas A. Keenan 
- Computer Systems Design Program, John R. Lehmann 
- Networking for Science Program, Walter A. Sedelow 
- Techniques and Systems Program, Sally Y. Sedelow 
- Software Quality Research Program, J. Richard Phil1 ips 
- Special Projects Program, Fred Weingarten 
Cmunications should be addressed to the Computer Science Section, National Sdence 
Foundation, Washington, D. C. 20550, or assistance may be obtained through the AFIPS 
Washington Office. 
FCC APPROVES WESTERN UNION REQUEST TO PROVIDE PACKET-SWITCHED SERVICE 
The FCC has approved a request from Western Union International to provide a packet- 
switched service between the United States and the United Kingdom, for the Advanced 
Research Project Agency (ARPA) in the Oepartment of Defense. The service is experi- 
mental in nature, and will be the first transatlantic application of packet-switch- 
ing technology. It will be used as a basis for gaining experience in transmitting 
between two packet networks (the second network being the Experimental Packet-Switch- 
ed System in the United Kingdom), and to explore the use of satellites in interna- 
tional networks. 
NBS A~NOUNCES NEW COBOL STANOARD 
The Repartment of Comerce has approved a new Federal Information Processing Standard 
(FIPS) 22-1, which makes Cobol-74 the new language standard applicable to all Federal 
agencies. Cobol is the only programing language for which a Federal standard has 
been established. It is maintained on a voluntary basis, and is presently used by 
8 government agencies. The new standard will require Cobol compliers used by the 
Federal government to comply with the 1974 Ansi Cobol specifications, and vendors 
will be required to categorize their compilers within one of four NBS levels, each 
of which consists of a specific set of Ansi "functional processing modules." 
PROGRAM ERROR GENERATES MASSfVE SOCIAL SECURITY OVERPAYMENT 
The cmissioner of the Social Security Administration has testified before an over- 
sight comittee of the House Ways and Means Comittee that Social Security overpay- 
ments of $461 million have been identified, and that the actual total may be as high 
as $800 million. The errors were made largely or entirely through the conlputer 
system which supports the Supplemental Security Income (SSI) program. Changes in a 
Complex statutory scheme have caused a haphazard development of SSI software; but the 
agency also admits to a lack of internal control, as well as a lack of coordination 
between systems or between programers working on the same system. The Comnissioner 
testified that many of the erroneous transactions are "literally beyond anybody's 
direct control." The errors have also led to concern that the system is vulnerable 
to fraudulent abuse by agency employees. 
PRIVACY PROTECTION STUDY COMMISSION MEETS IN WASHINGTON 
The Privacy Protection Study Comissioh met in Washington on September 8, with an 
agenda which included presentations From Federal government groups and prospective 
contractors. In response to a suggestion by chairman David Linowes, the Comnission 
resolved to create a conittee on Freedom of Information and Privacy, chaired by 
Minnesota State Sen. Robert Tennessen. Executive director Carole Parsons enumerated 
a list of areas possibly appropriate for the Comission's attention, which included: 
mailing lists; private sector usage of universal identifiers; valuntary private sec- 
tor compliance with Federal regulations; data collection criteria; credit card and 
reservation systems; Freedom of Information Act litigation; health and medical re- 
cords; multi-jurisdictional data systems; consumer reporting services; credit issu- 
ance and insurance; statistics and research; social services; employment and person- 
nel matters; oversight, enforcement and remedies regarding privacy policies; inter- 
national implications (including mu1 ti-national corporations) ; cost factors in imp1 e- 
menting privacy safeguards; and the Federal-state relationship in privacy regulation. 
The Comnission directed Parsons to initially look into credit cards, mailing lists, 
universal identifiers and the Federal-state relationship. 
TUNNEY HEARINGS ON WHITE HOUSE COMPUTERS 
The Senate Subcornnittee on Constitutional Rights, chaired by Sen. John Tunney, 
held fact-finding hearings on September 9 to investigate computer usage (pertaining 
to personal data files) by the White House and'the Federal Preparedness Agency (FPA). 
Tunney indicated that the investigation could lead to legislation controlling the 
use of surveillance technology. Witnesses for the White House Office of Management 
and Budget, and FPA, testified that their computers were not capable of direct com- 
munications with other agency computers, and that there had been no improper use of 
their systems. Tunney nonetheless expressed concern regarding accountability and 
control in the use of such systems. In particular, he pointed to the lack of knowl- 
edge on the part of White House witnesses concerning the data base content in the 
FPA computers at Mt. Weather, while the director of that system had indicated that 
its data base content was a matter for the discretion of other agencies. 
NEWS BRIEFS 
The Comnerce Department has announced that U.S. Exports of computers and related 
equipment totalled $555 million during the first quarter of this year; imports 
for the same period totalled $32 million. 
The White House Office of Telecmunications Policy has released a privacy study 
which evaluates privacy safeguards ana proviues recmendations. 
Brig. Gen. Walter E. Adams has been named Comnander, U. 5. Army Computer Systems 
Comnand at Ft. Belvoir, Virginia; succeeding him as Deputy Cmander will be 
men. Jack L. Hancock, formerly of the Defense Comnunications Agency. 
The U.S. Comptroller General has reconended that the Navy take measures in several 
areas to improve management of its data processing programs. 
The General Services Administration has solicited (RFP-CDPA-76-1) ro osals to 
develop ADP mdnagement guidelines for Government executives a* 
AFIPS IN WASHINGTON 
PRIVACY 
As a result of interactions to date, the Washington Office has received specific sug- 
gestions from bath the Privacy Protection Study Comnission and the White House Danes- 
tic Council Privacy Comnittee regarding assistance which AFIPS might provide. Our 
next activity will be to structure a group of individuals from the AFIPS societies 
who are willing to make a comnitment to providing this assistance, 
FEDERAL COMMUNICATIONS COMMISSION BRIEFINGS 
The Washington Office has instituted a series of briefings to the Federal Cmica- 
tlons Cmission, beginning with ivformal presentatioils by AFIPS experts visiting 
Washington. A near-term objective in this area is to provide a more comprehensive 
briefing to the Coninission, describing broadly how technological developments in, 
computing 'will affect communications and ultimately affect the regulatory responsi- 
bilities of that agency. 
FEDERAL COMPUTER NETWORKS AND PERSONAL DATA FILES 
The use or potential use of Federal computer networks to exchange personal data 
files has recently been the subject of attention by both Sen, Tunney's Subcommittee 
on~Constitu$ional Rights, and NBC Tews. 
Earlier this year, Ford Rowan of NBC pro- 
duced a series of national news stories based on his investigation of Federal net- 
works, particularly those in the defense and intelligence comnunities. With the 
objective of providing appropriate input from the computing profession, the AFIPS 
Washington Off~ce arranged for Mr. Rowan to interview Prof. Bernard Galler and Dr. 
Willis Ware; these interviews are expected to dppear on the NBC Nightly News. 
GENERAL INFORMATIONAL ACTIVITIES 
In addit-ion to producing the AFIPS Washington Report, recent activities have includ- 
ed: the development of a mechanism to disseminate NSF program solicitations froni 
both the Computer Science Section and the Office of Science and Information Service; 
providing infomation in response to individual requests from AFIPS constituents; and 
providing direct briefings to the AEDS and IEEE Computer Society boards which have re- 
cently met in Washington. In providing information to the Federal government, the 
Washington Office has recently responded to requests from the Federal Comnunications 
Comnission, the Library of Congress, and the Domestic Council Privacy Comnittee. 
PXPLORATORY ACTIVITIES 
An Important concept in the Washington Office charter is that the Office will func- 
tion as an early warning mechanism with regard to developments in the Federal gov- 
ernment. The objective is to permit AFIPS societies the opportunity to participate 
in such developments, rather than responding to them as faits accom lis. Accordingly, 
the Washington Office devotes substantial resources to ads hearings 
and meetings, and scanning on a daily basis such publications as the Federal Regis- 
ter, the Congressional Record and various comnercial reporting cervices. 
- 
Recent exploratory meetings have been held with several groups which have indicated 
a definite interest in access to technical information through AFIPS, e.g., the 
Congressional Office of Technology Assessment, the Air Force Systems Comnand, and 
the Office of Automated Data Processing Management in the General Services Adminis- 
tration. Continued liaison with congressional comnittees has permitted the Washington 
Office to alert AFIPS to major anticipated hearings, such as those to be held this 
fall by the House Committee on Science and Technology on (1) technology transfer and 
(2) the R & D programs of the National Science Foundation (inc1udin.g its program in 
computer resear h : . 
AFIPS societierriav-e permission to use inaterial In the AFIPS Washington RepoPt for 
their own publicdt~ons, except that where an article title appears with an (*) 
clearance must first be obtained from the AFIPS Washington Office. 
- 
d! Washington Report 
...................................... 
Amencan Federalion ol Inlormallon F'tmSung Sot elks Inr Su'lt* 37,) ?I00 L Slreel N W Wdsh'nplun P ?0113i 207 ?96.059(1 
Vol. I, No. 4 November, 1975 
WASHINGTON DEVELOPMENTS 
NATIONAL COMlISSION ON LIBRARIES AND INFORMATION SCIENCE ANNOUNCES GOALS 
The National Commission on Libraries and Information Science (NCLIS) ha$ stated 
its long range program in a docment recently transmitted to the House Conunittee 
on Education and Labor. 
NcLI.9. The Comission, established by statute in 1970, is chartered to advise 
the President and the Congress on national policy related to library and infor- 
mation services; its major program objectives are (1) to develop resources sup- 
portive of high quality library and information services, and (2) to form a 
national network of library and information facilities. Specific objectives 
are to " k]ncourage the private sector . . . to become an active partner in the 
development of the National Program," as well as to ensure adequate library and 
infomation services, provide adequate special services, strengthen state re- 
sources, ensure education of related personnel, coordinate Federal programs, and 
establish a locus of Federal responsibility for the National Program. 
Nationwide netwrk. The Commission envisions a nationwide network of libraries 
and infomation centers encompassing state networks, multistate networks, and 
specialized netwrks in the public and private sectors; the definition of "net- 
work" in this cmntext includes computers and computer communications, although 
it extend3 to cmventional telephone and teletype devices as well. 
While the Federal government would neither directly control nor operate the net- 
work, it would have responsibility for aiding the development of compatible net- 
works, furthering comn practices, and increasing coordination between the pri- 
vate and public sectors; it would also "collaborate with appropriate professional 
societies in promulgating interstate technical standards, support the introduction 
of additional coyuter and telecommunications facilities as needed for interstate 
purposes, and help establish protocols governing the way transactions are handled 
by the network." 
Other Federal (as opposed to private) responsibilities identified by the Comis- 
sion include developing centralized services for networking, exploring computer 
use, applying new forns of telecommunications, and supporting research and devel 
opment. 
Copies of the present National Program Document (U) may be obtained through the 
MIPS Washington Office; comments are, as noted above, encouraged by NCLIB. 
69 
FEDERAL PRIVACY ACT IMPLEMENTATION 
Ihe Privacy Act of 1974 took effect on September 27, amidst certain unresolved 
questions as to its operation. 
lhe Department of Health, Education and Welfare has characterized itself as a 
single agency for purposes of the Act, allowing (barring further Congressio- 
nal response) the 11 separate agencies within HEW to exchange personal data files 
without the disclosure records required for transfers between other Federal agen- 
cies. 
In a separate development, members of Congress expressed concern over the fact 
that sone agencies were interpreting the Act to preclude Congressional access to 
agency files of personal data. Members of Congress frequently request such 
access to aid their constituents in dealing with Federal agencies. 
The situation 
was resolved, temporarily at least, by an OMB memo recommending that each agency 
establish Congressional access as a "routine use" of a system of records under 
the Act; the effect on the OMB procedure will be to obviate the need for niembcrs 
of Congress to obtain written consent in order to access a constituent's data 
file. 
Finally, Sen. Kennedy has taken :I strong issue with a Justice Department opinion 
which operated to exempt records which an agency had exempted from public access 
under the Privacy Act, from access under the Freedom of Information Act (FOIA) 
as well. In response to Kennedy's objection, the Justice Department has proposed 
a new regulation which would allow the FOIA to apply to records exempted from the 
Privacy Act, but only at an agency's discretion. 
POLITICS AND PRIVACY 
A number of prominent politicians have been addressing the privacy issue recently, 
most notably President Ford. In an address ((1 at Stanford Law School, Ford spoke of 
the need to balance the individual right of privacy with the government's legiti- 
mate need for information related to its program. While Ford made clear his in- 
tolerance of illegal invasions of privacy, his coments were primarily directed 
toward threats to privacy which result from otherwise beneficial programs; in 
such areas, he said, 'ke must protect every individual from excessive and unnec- 
essary intrusions by a Big Brother bureaucracy." In a separate statement Ford 
later said that "N eeping only an essential minimum of these records is the most 
effective protection we have" against incursions on privacy. 
Sen. Tunney, chairman of the Subcommittee on Constitutional Rights, recently re- 
affirmad his concern with privacy in a speech (#) on the Senate floor. He noted 
that his Subcommittee was presently moving bills which: would prohibit military 
spying on civilians; would provide for control of criminal records (Washington 
Report, 9/75); and would secure greater privacy protection for government employ- 
ees. Turtney pointed out the need to anticipate technological threats to privacy, 
an endeavor in which AFIPS has already offered assistance to the Senator. 
Speaking in a recent interview in EDP Weekly, Rep. Goldwater stated that he hoped 
for an attitude of support and voluntary compliance with privacy protection prin- 
ciples in the colmPercia1 sector. While he feels that the business sector has an 
incentive to be concerned with privacy as a "good business practice," he feels 
that a similar incentive does not exist in the area of government; he pointed out, 
however, that he expects the states to enact their own privacy legislation, and 
that the Privacy Commission Cof which he is a member) would be looking primarily 
at the Federal-state relationship. In a speech at the recent IEEE Computer 
Society CWCON, Goldwater also urged the information processing community to 
provide technical input as a basis for reasonable legislation. 
CONTROVERSY ON,DEFENSE IR$D PROGRAM 
?he Department of Defense Independent Research and Development IIRl,ll) progr~mr 
has been the subject of controversy in hearings before the Senate Irmed Services 
Subcommittee. The program permits defense contractors to charge part of their 
general R 6 D programs as overhead-on defense contracts; its cost in 1975 ros 
over $800 million dollars. 
Proponents of the program argued that it allows contractors to keep abreast of 
current technological developments, and that it allows the military to buy .td- 
vnnCed electronic system at a lesser cost than if developed under n. Jircctly 
funded R 6 D program. Critics of the program, however, testified that ~t allow:. 
firms to charge part of the cost of developing commercial projects to their mili- 
tary contracts; Admiral Hyman Rickover, heatl of the Navy Nuclear Propulsion 
Program, recornended that the program be eliminated. 
COMPTROLLER OF CURRENCY EFTS RULING ENJOINED FROM IFIPLEMENTATION 
The Comptroller of the Currency has suspended his December, 1974 ruling (#) that 
Customer-Bank Comunication Terminals (CBCT's) are not branch banks within the 
meanipg of the laws governing national banks. A decision last July in the Federal 
District Court in Washington (Washington Report, 9/75) had enjoined fuxther imple- 
mentation of the ruling, but it remained in effect while the Comptroller requested 
a stay of the District Court decision, pending appeal; on Octaber 10 the stay was 
denied and the ruling immediately suspended. 
PRESIDENT APPQIWS EFTS CCMdISSION 
President Ford has appointed the non-statutory members of the EFTS Commission, 
over one year after the enactment of the statute (t) creating the Commission. As 
a result of the delay the Commission will have less than a year to complete its 
work, unless the Congress passes an amendment (presently pending) which would 
extend its period of activity. William B. Widnall, former Congressman and ranking 
Republican on both the House Banking and Currency Committee and the Joint Economic 
Committee, has baen nominated as chairman; the Senate Banking Committee held brief 
confirmation hearings on October 23, and confirmation was expected (at press time) 
by Novembek 1. 
Members representing the financial sector are Richard D. Hill (First National Bank 
of Boston), Verne S. Atwater (Central Savings Bank of New York), and R~~ G, G~~~~ 
(Fidelity Federal S & L). State banking officials appointed were James E. Faris 
(Indiana) and William B. Lewis (New Jersey), and Herb Wegner was appointed as a 
trade association executive. Other nominees were Freyda P. Koplow, (former Massa- 
chusetts Commissioner of Banks), Fairfax Leary, Jr. (Temple University law profes- 
sor); Almarin Phillips (U. of Pennsylvania) ; Gordon R. Worley (Montgomery Ward) ; 
George W. Waters (American Express); John J. Reynolds (Interbank Card Association); 
and Ralph F. Lewis (Harvard Business Review). 
7 1 
NSP ACTIVITIES 
mt Events. The National Science Foundation held an October 28 conference at 
Case Western Reserve University on Production Research and Industrial Automation; 
part of Vhe NSF Research Applied to National Needs Program, the conference focused 
on technological improvement of productivity. 
NSF also has awarded a grant to 
John W. Hamblen, chairman of the University of Missouri Computer Science Depart- 
ment, to conduct an Inventory of Computer Acrivities and Related Ilegree Programs 
in U.S. Higher Education. 
Nnding Opprtunities. 
NSF is seeking proposals from U.S. institutions, for five 
day mgional conferences, on subjects of current research interest in the mathemati- 
cal sciences. The Foundation has also reopened its annual competition for faculty 
fellbwships in Science Applied to Societal Problems; awards will be offered in all 
scientific fields, including computer science and mathematics, and applications (#) 
are due by February 6, 1976. 
ASSISTANT SECRETARY OF CWRCE ON COMPUTER-AIDED MANUFACTURING 
Speaking at the recent Second International Computer-Aided Manufacturing (CAM) 
Standards workshop in Washington, Dr. Betsy Ancker-Johnson, Assistant Secretary 
of Commerce for Science and Technology, described the potential of computer-aided 
manufacturing for increasing national productivity. In a speech reported by EDP 
Weekly, she stressed the opportunities related to the standardization of CAM 
Systems, pointing out that efficient applications will depend on standards in data 
base formats, computer languages and interfaces between system components. 
DATA COhMRJICATIONS ISSUES AT THE FCC 
Telenet and Tymshare. Telenet has filed a request with the FCC, seeking a ruling 
that meshare be required to operate as a regulated common carrier, alleging 
that meshare's policies do not conform with FCC rules on computer-based commu- 
nication services. While both companies provide packet-switched network services, 
only Telenet has filed with the Commission as a common carrier. Regulation of 
such value-added networks is also an issue in the current FCC "Resale. and Sharing 
Docket;" this inquiry (X) is investigating whether non-regulated carriers which 
purchase cmication services at bulk discqunts (and subsequently resell or 
share them) should be treated as common carriers. 
Aetna joins cML satellite. IBM, Comsat General, and Aetna Life E Casualty have 
proposed ta the FCC a partnership arrangement in which each would become an equal 
owner in ML Satellite, a corporation being organized to provide satellite data 
comunications; the proposal corresponds to the "balanced option" for CML, approved 
by the FCC in an earlier ruling. The earlier FCC decision allowing an IBM-Comsat 
General joint venture remains the subject of litigation by three existing satellite 
carriers, Western Union, RCA Global Counnunications, and American Satellite. 
NEWS BRIEFS 
The National Academy of Sciences has published a report (#) recommending establish- 
ment of a National Resource for Computation in Chemistry. 
7 2 
Lear Siegler has agreed to an FTC consent decree, based on allegations of unfair 
and deceptive sales practices, requiring refund of computer course tuitions 
to eligible former students. 
The National Bureau of Standards has received a National Science Foundation grant 
to study the impact of copyright and other laws on the economics of technol- 
ogy related to scientific and technical information systems. 
The White House Office of Telecommunications Policy (OTP) has fomd a committee 
to consider long-range export policy issues; the group will br coordinated 
by the OTP Assistant Director for International Communicationb, William 1. 
Fishman. 
President Ford has nominated Charles Slichter to the Ndtional Sclence Board of 
the National Science Foundation. 
The National Academy of Sciences has published its current report [#) of Fellw- 
ship and Research Opportunities in the Matbematical Sciences. 
The General Accounting Office has released reports which (1) criticize the Office 
of Management and Budget for not adequately centralizing computer procurement, 
(2) recommend improved management of information systems detelopment in the 
Federal Aviation Administration, and (3) recommend cancelling procurenlent of 
a major Department of Agriculture system, due to inadequate agency planning. 
Complete sets of the Federal mfomtion processing staodards (FIPS) are now avail- 
able from the National Bureau of Standards at $46.00 each. 
The Computer and Business Equipment Manufacturers Association has published a 
series of brochures on privacy (#), discussing issues and presenting state- 
ments by major spokesman within, and outside the Federal government. 
AFIPS IN WASHINGTON 
MIPS BRIEFS DOMESTIC COUNCIL PRIVACY COMMITTEE DIRECTOR 
On October 3rd, representatives of AFIPS societies provided a briefing to Qulncy 
Rodgers, executive director of the White House Domestic Council Privacy Committee. 
The briefing to Rodgers was primarily in anticipation of a vice-presidential round- 
table on Privacy and Information Policy. AFIPS participants were: Daniel McCracken 
(ACM), Charles Meadow (ASIS), Bill Moser (DPMA), and Bruce Peters (IEEE Computer 
Society). 
The objective of the briefing was to provide Rodgers an assessment, from the 
prospective of professionals in the information processing field, of issues relat- 
ed to Federal government information policy. Vice President Rockefeller, chairman 
of the Domestic Council Privacy Committee, 1s concerned with the development of a 
comprehensive policy which would view these Issues in a broad perspective, in 
contrast to the ad hoc fashion in which they are presently handled by the Federal 
government. 
COORDINATING AFIPS COWITTEES AND WASHINGTON ACTIVITIES 
Since several existing MIPS committees address areas of continuing interest in 
Washington, an effort is undexway to closely coordinate such committees with the 
activities of the AFIPS Washington Office and to make the facilities of the office 
available in connection with relevant committee work. AFIPS committees initially 
identified for close liaison include those on Privacy, EFTS and Statistics. 
AFIPS NASHINGTON ACTIVITIES CWIITEE MEETS IN WASHINGTON 
An MIPS Washington Activities Committee meeting was held In Washington on October 
8, attended by MIPS Presiden Tony Ralston and Executive Director Robert Rector. 
In additian to conducting its regular review and assessment of Washington Office 
activities, the Connnittee resolved to pursue specific projects with the Federal 
Communications Commission [FCC) and the Institute for Computer Sciences and Tcch- 
nology (ICST) at the National Bureau of Standards; both of these projects are 
presently in an exploratory phase. The objective of the FCC project is a major 
briefing to the Commission, assessing the impact of developments ~n computer 
coi!ninmications on FCC regulation and the ICST project is oriented toward 
an 
assessment of the scope, character and impact of computer scienc~ as a discipline. 
N6W WASHINGTON- RGPORT POLICIES 
At its recent meeting, the Washington Activities Committee established three major 
policies for the AFIPS Washington Report: 
Subscriptions to the Washington Report shall be, for the present, 
available upon request. While the publication is viewed primarily 
as a member service to MIPS societies, it is available to non- 
AFIPS groups and individuals as well. Ib is anticipated that sub- 
scription charges will be imposed at a future date, with approprl- 
ate differentials for AFIPS societies and society members. 
Key documents discussed or cited in the Washlnyton Report will be, 
made available through the Washington Office upon request, whcre 
feasible; availability will be indicated by a "(#)'I symbol in the 
text of the Washington Report. 
In appropriate instances, the Washington Report will solicit comments 
and provide a form for discussion of particular events or programs in 
the Federal government. 
In addition to the foregoing services, AFIPS societies can, of course, contact the 
Washington Office for assistance in connecti~n with any matters discussed in the 
Washington Report; it is indeed hoped that arhicles will precipitate both .inquiries 
and comment. 
ZSPUWS WHITE HOUSE BRIEFING 
A joint effort by AFIPS Executive Director Robert Rector and the AFLPS Washington 
Office has resulted in tentative plans for a February 5, 1976 White House brief- 
ing to senior officials of the AFIPS societies. AFIPS will be brieted by various 
executive agency spokesmen in areas of particular interest to the ~nfonation 
processing field; prospective topics include research and development, technglogy 
transfer, EFTS, computer communications, privacy, Federal usage of computers: and 
standards. In addition. AFIPS may undertake to brief ?he Administration on 
issues as viewed by professionals in infonnation processing. Comments on proposed 
topics are solicited from the AFIPS socleties, 
AFIPS societies have permission ta use material in the AFIPS Wasl~ington Report fo~ 
their own publications, except chat where an article title appears with an " * 'I 
clearance must first be obtained from the AFIPS Washington Office. 
Documents 
indicated by the symbol "(B)" are available on request to the AFIPS Washington Office. 
d! Washington Report 74 
...................................... 
Ami.nr,~n Fudrrduon ol lnlormdllon Pruwssng Soc~eltes Inr Su~le 470 2100 LStreel N W WdShlngIOn. DC 20037 202-296 0580 
Vol. I, No. 5, December, 1975 
WASHINGTON DEVELOPMENTS 
FCC APPROVES DIRECT INTERCONNECTION OF TERMINAL EQUIPMENT 
The Federal Communications Commission (FCC) has ruled that users may directly 
connect terminal equipment to the telephone network without the use of carrier- 
supplied connecting arrangements, provided they comply with FCC requirements. 
The ruling (#) is to become effective April 1, 1976. 
Impact of the decision. The ruling will permit users to connect computer termi- 
nal equipment to the network without the necessity of either a modem supplied by 
the common carrier (generally ATFjT) or a telephone company data access arrange- 
ment (DAA); the DM had previously been required if a Bell modem were not used. 
While circuitry similar to the DAA will still have to be incorporated in customer- 
supplied equipment, it is expected to be considerably less expensive to users 
than present AT&T lease charges. 
Xistory. The FCC ruling requiring the use of telephone company connecting ar- 
rangements for user-supplied terminals has been the subject of controversey ever 
since its implement?tioh after the Carterfone decision (which allowed the connec- 
tion of custo,mer-provided equipment) in 1968. The common carriers have argued 
that such arrhghents are necessary to protect the network against potential 
hams, such as those identified by a 1970 National Academy of Sciences study: 
hazardous valtages, excessive signal power levels, excessive longitudinal imbal- 
ance, and imp,roper network control signaling. Opponents maintain, however, that 
the telephone cmpanies have insisted on carrier-supplied access arrangements to 
protect their own revenues. 
Registration program. The FCC will allow interconnection only if it is done 
through either protective circuitry registered with the FCC, or directly through 
equipment which is itself registered. The Commission has also required that 
carrier-supplied terminal equipnient be registered (a requirement not generally 
imposed on comon carriers) in order to enhance network protection, insure compet- 
itive equality (between carriers and non-carriers) in the manufacture of such 
equipment, and to prov'i'de a "benchmark" against which to judge non-carrier 
applications. 
Possible appeal. ATBT is expected to appeal the FCC ruling, which may delay its 
effective date. Further, the ruling applies only to the technical aspects of 
intercomection; a separate proceeding (Docket 20003) is underway to examine the 
potential economic harm from competition in this area. A reversal of the technical 
ruling would appear unlikely on economic grounds, since data communications devices 
account for only the small portion of ATET revenues. 
7 5 
OTP REPORT ON SOCIAL ISSUES IN EPTS 
The White House Office of Telecommunications Policy (OTY) has relensed a study 
on the non-economic implications of EFTS, entitled "Value Choices in Electronic 
Funds Trnnsfer Policy." In transmitting the study to the Vice President, OTP 
Acting Director John Eger said "both because of our particular mandate to evolve 
policy relating to the intorconncction of contputers with teleconununications, an 
intcrconncction which is tlie necessary foundation of any EFT system, and because 
of our concern for i,.formed and cffcctive formnulution of policy in the executive 
brdnch, DTP has begun to scrutinize closely the increasing range of Fedcral gov- 
ernmcnt activity in this area. . . . While mnny of the aost significant economic 
questions have becn rarscd, critical non-ecunomic issues have not bccn addressed- 
such as the desirability of unrcttered govcfnltrent acccss to pcrsonnl records ei- 
ther as a user or an operator of alty EFT system." A prim:~ry signific:lnce of the 
rcport is thc indication that OTP is squarely concc~ncd irith the social in~p:ict of 
EFTS. This subjctt is presently comlsidcrcd (or pl;~nncd to I)c considered) only 
peripherally by groups such ns the I'rivacy Cummission, the IiF'l'S Cot~~mnission, or 
the Fedcral Co~~~i~~nications Cu~~unission. 
IVIIITE llOUSE SCIENCE AIIVISER R1I.L PASSES HOUSE 
--- -- 
A new bill to cst.tblish a prcsidcntial scicnce adv~ser, 1i.R. 10230, has bccn 
passed (362-28) in the Ilo~~se of Reprcsentntivcs. The lcgislntion, draftcd coop- 
er~tivcly by the Ford Admninistrntion and llousc scicnce lc.idcrs, is a revision of 
H.R. 9058 (Washinyton Report, 9/75). 
In its present form, the bill (1) cstdblishcs-a nntlon:~l polxcy for scicncc and 
technology, (2) establishes an Office of Scicnce 4 Technology Policy (OSI'P) in 
the White House, and (3) establishes a Federal Scicnce 6 l'cclinology Survcy Conunit- 
toe (the Survey Committee) in thc White IIouse. The Survey Comlu~iittce rill have two 
years to analyze Federal science and technology cfforts .ind lcport its findings to 
the President, trho must subsequently review and transnut the report to Congress 
with his recommendations. The director OF OSI'P will scrve as the President's Scl- 
ence Adviser, and as chairman of the Survey Co~imittce. 
Despite the likelihood of this bill becoming l~w, thcrc rca1.1ins serious concern in 
the scicnce commln~unity (c.g., llillinm C:lrey's editorial in thc Koven~ber 21 issue of 
Science) that science budgets may sufrcr st~bst:~mitinl cutb,~chs in Prosident Ford's 
program to reduce Federal cxpcndituncs by $28 million. 
NEW NSF ADVISORY GROUPS 
President Ford has created two new science advisory groups withln the National 
Science Fo,undation, to advise him on planning for tlie proposed Office of Science 4 
Technology Policy (see above surmnary of H.R. 10230). These arc; respectively, the 
Anticipated Advances in Science 4 Technology Advisory Group, and the Contributions 
of Technology to Economic Strength Advisory Group. The former group, headed by 
Dr. William 0. Baker (president, Bell Labs) will advise the President on national 
policy implications of developments in science and engineering; the latter group, 
headed by Dr. Simon Ramo (board vice chairman, TRN) will advise on imitproving the 
utilization of technology to foster economic strength. 
In additian to the advisory group chairmen, other members related to the con~puting 
field include Lew Branscomb (Vice President and Chief Scientist, IBM), Joseph Charyk, 
(president, Comsat) and Patrick Haggerty, (board chairman, Texas Instruments). 
7 6 
PRIVACY COMMISSION ACTIVITIES 
The Commission is presently investigating eight topical areas. It has thus far 
produced staff reports in four of these, "The Use of Mailing Lists in the Private 
Sector," "The Use of the Social Security Number in the Private Soctor," "Disclo- 
sure of Federal Income 'Tax Returns to Third Parties,ll and "Credit Card Kccord- 
Keeping: The InFormational Privacy Issues," and it has inunediatc plans to look in- 
to the remaining areas: consumer credit reporting, en~ployment records, social 
services, and statistical research. The Commission presently receives input from 
its own staff (which will be supplemented by contractual support of about $50 
thousand per year), industrial groups and non-profit organizations such as AFIPS. 
The Commissioners rcccntly indicated that they will not postponc their rccommenda- 
tion until their final report is due (June, 1977), but will publish proposed rcc- 
ommendations for public commcnt as the Commission proceeds. 
NEWS BRIEFS 
The NSF Advisory Panel for Computer Science & Engineering met recently to discuss 
NSP programs in Softwarc and Progra~ning Systems, Software Quality Research, 
and the general program in computer science. 
The Federal Home Loan Bank Board, in response to pressure from the Justice Depart- 
ment, has agreed to abstain from advocating government-sponsored EFT, and to 
support private sector development of EFTS systems. 
The Association of Data Processing Service Organizations (hDnPSO) has flled a 
brief in the Johnston software patent case before the Supreme Court (Wash~ngton 
Report, 9/75) supporting software patentability, and opposing the brief filed 
earlier by the Computer 4 Business Equipment Manufacturers Association [CBEbLk). 
The National Science Foundation is seeking proposals for participation In U.S.- 
U.S.S.R. activities in the application of computers to management; present 
joint activities are (1) econolnetrlc modeling, (2) computer analys~s of eco- 
nomics in management of large systems, (3) applications of computers to man- 
agement of large citics, (4) theoretical foundations for software applica- 
tions in econo~nics in management, and (5) computer-aided refinement of deci- 
sion making, and education of high-level executives. 
The Stanford Research Institute has received a $275,000.00 grant from the Experi- 
mental Technology Incentive Program in the National Bureau of Standards, to 
develop guidelines for improving research and development planning in fifteen 
non-defense Federal agencies. 
The House Committee on Science & Technology has published its Annual Report on 
Federal R & D Programs (#) . 
The Law Enforcement Assistance Administration has published a compendium of-state 
laws which govern security and confidentiality of criminal justice informa- 
tion systems. 
Attorney General Levi has denied permission to the FBI to proceed with the 
development of its computerized message-switching system; pending passage 
of further legislation. 
AFIPS IN 1VASIIINGTON 
AFIPS PROVIDES CONGRESSIONAL TESTIMONY ON TECIWOLOGY TRANSFER 
- 
At tho request of the llouse Committee on Science 4 Technology, Sul)committce on 
Domestic and International Scientific Planning and Analysis, AFIPS provided wit- 
ncsses to tostify at reccnt hearings on technology transfer to nations in the 
Organization of Petroleum Exporting Countries (OPEC). APIPS provided two wit- 
nesses, Prof. J.C.R. Licklider of MIT and Prof. Harry Huskey of the University of 
California at Santa Cruz, who testified specifically on t~ansfcr of computer tech- 
nology. 
Dr. Licklidor distinguished in his remdrks eight aspects of technology subject to 
export: hardwars, softworo, systcms, networks, services, inforn~ation, education 
and individual expertise. Ilo urged that while computer scrviccs and accqss to 
computing resources may be appropriate for export, that exports of computer hard- 
ware should be scrutinized for their national security implications. Prof. Iiuskcy 
pointcd out that information technology cnn be most hclpful to developing notions, 
and tends to have a stsblifing cffcct on these countries; he citcd the cxdmplc of 
Management Information Systems in food distribution. Ile stated further that en- 
couraging utilization of the technology by developing nations is beneficial to the 
U.S. economy andstimulates do~~lcstic R F, D (which in turn maintains our relative 
technological advantage). Trsnscripts of the hearings (#) will be published in 
about one month. 
PRIVACY COhhlISSION REQUESTS AFIPS EXPERTISE 
The Privacy Protection Study Commission has lnede a request of AFIPS to provide a 
panel of experts familiar with certain aspects of the use of the Social Security 
Number (SSN) as a universal identifier. The Commission is interested in conuauni- 
cating with thrce specific categories of individuals: 
(1) managers of record kceping systems who are familiar with current 
prlvate sector usage of the SSN, 
(2) persons directly familiar with an operation which has recently 
converted (perhaps in anticipation of private sector legislation) from 
the SSN to some other numbering system, and 
(3) individuals in state and local goverments familiar with the impact 
of Section VII of the Privacy Act, which makes it unlawful for such 
governments to deny benefits to individuals refusing to disclose their SSN. 
In response to this request, the APIPS lashington Office has contacted the liaison 
from each AFIPS society asking them to provide nominees to the AFIPS panel. 
AFIPS societies have pennission to use material in the AFIPS Washington Report for 
their own publications, except that where an article title appears with an " * " 
clearance must first be obtained from the AFIPS IVashington Office. 
Documents 
indicated by the symbol 'I(#)" are available on rquest to the AFIPS Washington Office. 
d* Washington Report 
7 8 
Vol. 11, No. 1 ,: 1!11!,1ry, l!1-11 
NATIONAL ACADEMY OF SCIENCES FORMS AI) llOC COMPUTER SCIENCI; PMNINI; (;RO!iI! 
The Assembly of Mathematical and Physical Sciences (MIPS) in thc National \c.~Jriny 
of Sciences (NAS) has recently formed an ad hoc ComputerSciences l'lanning Group 
to consider activities which AMPS might undertake in that field. 'Ihc L;rauli is to 
consider "what aspects of this rapidly growing discipline and wh~t probie~~~s rel~t- 
ed to federal and non-federal applications might benefit through special studies, 
particularly by [WS] ." 
The AFIPS Washington Office has confirmed that the formation of the Comliutcr 
Science Planning Group was precipitated in part by letters earlier this year to 
Dr. Philip Handler (president, NAS) from Anthony Ralston and Jean Sammet, presi- 
dents of AFIPS and ACM, respectively. Both Ralston and Smet urged Handler to 
consider organizational changes which would recognize computer science as .I sepa- 
rate discipline within the National Research Council [NRC -- the NAS and the 
National Academy of Erlgineering), arguing that this recognition has already oc- 
curred within major universities and throughout most of the academic commun~ty. 
History. There have been a number of activities within the NRC in the last ten 
years which relate to computer science and engineering, most recently the Computer 
Science and Engineering Board (CSEB). The CSEB was in existence from 1968 to 1972, 
chartered to assess the implications of information performing technology with 
regard to the public and private sectors of the United States. The National RE- 
search Council generally operates by performing studies under contract to various 
Federal groups, and occasionally undertakes similar studies on its own initiative; 
there have been no studies primarily related to information processingsince the 
termination of the CSEB. 
While the Academies have approximately n dozen computer scientists and engineers 
as members, the computer science discipline has not been organizationally recog- 
nized outside of existing groups in mathematics and engineering. 
Meeting agenda. At its recent first meeting, the Computer Science Planning Group 
undertook a general discussion of areas which would be candidates for NRC atten- 
tion, including (1) the impact of computers on soclety, (2) computer engineering, 
(3) computer science, and (4) the commercial impact of computer technology. Re- 
ports will be produced in these areas by early 1976. 
Outlook for a permanent group While there is apparently no consensus in the 
Group on how to proceed, it appears highly likely that a recommendation will be 
forthcoming to set up a permanent computer science group within the NRC; the prin- 
cipal unresolved questions relate to the structure of such a group. There are 
presently three other groups proceeding in parallel with the AMPS group, each of 
which will ultimately produce a report recommending NRC activities in this field, 
these are the Assembly of Engineering, the Assembly of Behavorial and Social 
Sciences, and the Connnission on Sociutczhniciil Systems. 
Pla~ing grasp nembers. 
Members of thc ,\+ll'S group arc Gordon Bell (Ill.:(:), Ruth Davis 
(National Burewu of Standards), Bcrn:~rd G.II ler (U, of blichigan) , IN~cll.lrd Garwin (IBM) , 
Jerrier Haddad [IRM] 
Richard Kerp (U. of L,iliforniat, ,Joshua Lcdellicrg (Stanford 
Medical School), J.C.R. Licklidur (YIT), Allrn Xcwcll (C;~rncgic-bi~l inn), 1;cor~c Pake 
(Xerox), Alan Perlis (Yale) , I':lt rick Supp?.; (StanFord) , and Ivan SII. 11r1.l ,lnd (RAND) . 
WUI AND IT DATA COWNICATIONS \PI'LICA'l'IONS BBFORI: I:CC 
WUI. Western Union International (Wl) has received ;ipproval from th' Fcdcr:ll Com- 
munications Commission (FCC) to Furnish its Internationnl Digital ll;~t;~ Scrv~ce (1DI)S) 
to Prance, Italy, Spain a~d Austria. The scrvico is cl;~imcd to provide high quality 
and reliability, primarily by simultnneous transmission of data by undersca cable 
and satellite. Thc FCC approval is for one year of operation, within whicli IDDS 
must submit an analysis af its service. IDDS will be required to submit a t:~riff 
proposal before beginning service, based on transmission speed (from 50 hits to 9.6 
kilobits per second), transmission volumc and other f:ictors. 
ITT. ITT Domcstic Transmission Systems has filed an application with thc I:CC nsk- 
ing for approval of a domestic data commu~iicntions network. Thc proposed scrvice, 
to be known as Corn-Pak, is a packet-switched system which woul'd be cap.lble of ser- 
vicing incompatible computers and computer terminals. The network would initially 
serve 13 cities by 1977, eventually expnnding to 24 cities. Like other value-added 
carriers (e.g., Telenet) Com-Pak would rely on circuits leased frum common carriers. 
While computer-related transmissions on the network will generally be immeditite, a 
delayed-transmission service (at a lawer price) will also be available to customers 
for'applications such as word processing. 
Related issues. ITT World Comunications has bpposed the FCC authori:ation of IDDS, 
arguing that the FCC should conduct a techaicill inquiry on optimum methods for inter- 
national digitnl data services, before g~xnting the IDDS application. 
On a separate topic, Western Union and other carriers have petitioned the United 
States Court of Appeals in Washington to reverse FCC approval of the proposed CML 
data communications satellite venture (Washington Report, 11/75), asserting that it 
would have monopolistic effect (such as dividing the market for satellite data com- 
munications between ATET and CMI,). The petitioning firms feel the monopoly condi- 
tions which might exist subsequent to CML entry into the market could not be reversed 
by antitrust regulation or litigation. 
COMPTROLLER OF CURRENCY EFTS SURVEY 
While the EFTS Comission continues to organize, a recent survey released by 
Comptroller of the Currency, James E. Smith, indicates the extent and distribution 
of existing EFTS systems. According tb Smith: 
The survey, which had a response rate of 97 percent of 4,700 national 
banks, showed that fully 10 percent of the banks had at least one 
Automated Teller Machine. As expected, a high proportion of large 
banks have an EFT system - 72.9 percent ot billion dollar banks and 
48.4 percent of those in the half billion to billion dollar range. 
However, more than half of ;ill 1'1"l' cystcms .ire in banks with under 
$100 million in deposits. ,I third .ire In hanks wit11 less than $50 
million in deposits. (Interestingly, u~.h;uli::~tion is not the key 
indicator of which hanks wi 11 h.rrc rht- 111.1rll incs. W:ishington, Oregon, 
Virginia, Mississippi nnbl West Virginid r:unl. III the top ten states 
having tho highest proportion nf hank, kith 1.t'I systems. New YorR 
and I'cnnsylvnnia arc 33rd ,inJ 3Stl1 in thr Ilnion rc.;per'ti\~cly.) 
The .lu.;ticc Department hns urged thc I:cdrr:~l Ikc.rrrr Hoard to modify it.; policy an 
con~putc.ri:ed chcck-clearing t':icilitic-: (Auto~n,lt~>rl Clearing Il~iuses -- \('HI<! opcrat- 
cd liy I'cderal Reserve Bank. The Ilcp:~rt~ncnt III.II'I~.I~I~:: thilt thc Bo~rd'h proi)o.;~d 
rule to provide only Fcderal Rcscrvc S\.;tcnl m1~111l)i.r Ir:inLs with unrcstrirlc,l access 
tn A~:il's would effectively dirc~.i~nin;~:r ;~g,iillcr t~tin-mrml~cr hanAs on3 tilrit't institu- 
t ions (hovinga and loen nssocit~t ions, ;III~ crvd~ t unions) ; t'urtil'r, if iodicr1Ycd that 
if. ACllts ot'fercd co~~~prri~cnsire scr\'ici.s h ~thtlut i'll;~rgc (r.g., proccssin~ dchit items) 
privatr invcstn~rnt and innor..~tion ill I:I"I'S, ii.~rt ii'ul,~rly point-of-sale .\'.;rems, would 
be discouraged. 
The Justicc llepartmcnt reconuncndcd that tllc HL,.I~LI rcctrict \TtI ser\.lces to appllca- 
tions such as payroll deposit plans, and w;~.il security payment plans; it recom- 
mended against services that would compct'v ~pninst point-of-sale retail financial 
services. It also urged the Board to adopt ;I pricing structure for ACH services 
which would reflect the ful.1 cost of providing sltch services to each participat~ng 
finonci;ll institution. 
WItI'l'E HUUSII SCIENCE tUWISEK I)EVELOPMEWIS 
The bill to cstabli.sh a Wl~ite llouse scicnc' ;~rir~scr, pdsscd recently In the House 
(IJ~sh!nqton Report, 12/"51, has encountcrril sul,.;t.rnti:~l d~ffiiulty in the Senate. 
Wh~lc the Ford c\d~ninistmtion had urged Scn,~tc p;~ss:~pe of a bill essentiall!. ~dcntl- 
cal to the cnrlier House version, the Sm,~tr is i~luving toward a hill less desirable 
to the President in that it would (1) rcrluirc the scicncc r~dvlser to mohe an :lnnnal 
report to the Congress, thus making the I'rcsidrnt ~ccountohlc for his differences 
with the :~dr,iser, (3) give the science adviscr rcspilnsihility Sor Funding recornmen- 
dations (a task the I'msident would prefer to hare left to the chccutive agencies 
and OMB], dnd (5) gcner;llly give the President lcss flexihil~ty in structuring the 
office of the science .idviser. The Senate cimmlttee:; handling the bill are Labor 
and Public Welfare (Sen. Kennedy), Commerce (Sen. Tunne).), and ~\cronautical dnd 
Space Sciences (Sen. Moss). 
The two National Gc~ence Found.~t~on advisury groups fornlcd by the President to 
assist in plannt~~!: ior tile office of office of Silence and 'Technology Policy (Wash- 
ington Report, I? -:-) held their first meet 111s in A.~.;htngton in December, at which 
they heard hrieS111ys fro111 relcv.mt feder~l grulilis 11 :ltiding tllr khite House Office of 
Management and 811tl.:ct, the Departments nf <t,~tt- i'o~io~~.rcc, \griculture, and the 
Federal Energy 11,lm~ 11 istrution. Ihe gruull. ~\csc ;11zt> IlriefeJ by heads of the National 
Academies, the Ol i~:c of 'Technology .Ass?. i>~nt .II,J tilt3 \i.ltion;~l Sclence Foundation. 
Chairing the joint mccting, Simon llnmo c.11 IL~ ~,II. init in1 discussion of national 
science policy issues appropriate for IVllitc ll~>~~sc attention, and asked each of the 
advisory group members to submit at their neht nlreting .~pproximately two such issues 
for further consideration. The AFII'S W:~shington OfFice plans to provide input to 
the advisory groups in this issue idctitification process; timely coments frm the 
AFIPS societies (i.e., preferably by .January 1.3, hut otherwise as early as possible) 
are, of course, solicited. One advisor). group member has indicated a substantial 
interest in Federal regulatory reform (which, wit11 regard to the information process- 
ing fic,ld, could include such areas as computer L~ommunications, electronic fund 
transfer systems, privacy, technology transfer, III~ computer industry antitrust). 
NEWS BRIEFS 
The General Services Administration has rclc;l.;l%l its unnu:ll "Inventory of Automatic 
Data Processing Equipment in the U.S. r:~ii~~~rtimcnt" (#I, for fisz.~l year 1!175; 
while IBM still has the most inst.11lcd vq\~ipmcnt (35:.), DEC was the lending 
vendor in 1975. 
The National Sciencc Foundation has ow.~rdccl ovcl. $1 million in grants for studies 
on the impact of Federal refiul;~tory ape~~cich, including Lhc I:cdvr.~l 
Communications Commission. 
The National Science Foundation has issuhd :I solicitation for propos.~l. [b) for the 
Educatiolial Program Restructuring pmjiram, which is to "encourage tho dcvelop- 
ment, testing and evaluiltiun of new or ~lncunvcntional approarhrh to all 
aspects of science instruction at thr ~lndergr:~duatc level." 
Charles C. Joyce, former assistant director of thc White llousc Officr (>I' Tclerom- 
munications Policy (OTP) has heen n.ln~cd VTI' speci~l as.;istant for national 
security and emergency communications. 
The nominee for Secretary of the Air I'orcc, Thomas Reed, comes from a present post 
as director of Pentagon telecom~~nicntions and commc~nd :~nd control systcms. 
Requests for coments on a proposed Federal information processing -t.lndard (PII'S) 
for the National Communications System have been requested by .Innuor). 30, 
197b; details are avnilable from AFII'S W:ishington (Iffice. 
The OFfice of Sci-encc Informatio! ervicc in the National Scienc(. Poundation has 
released its Summary of Awards I'or fiscal ).c:lr ]!I75 (x). 
While the U:S. vs. AT&T antitrust suit is del:~yed in court, the Scnote ~\ntitrust 
Subcommittee is considering lcgls1.1tion to rcorganizc .\l'tT; o suggested 
bill has beon submitted by the Computcr Indt~strirs hssoci~ltio~i. 
Thc Federal government has denied an IBM request to export il 3iO/!SS-bascd 
reservations system to the Soviet Union; ;I primary factor 111 the decision 
was thc approximately thirty model 3330 disc drives involi'rJ in the systcm. 
The Coml~uter Industry Association has recomendcd stronger Ant 11 rust legislation 
in recent testimony relating to the I'rcdatory I'racticcs \rt of 1975, before 
the House Committee on Seal1 Business. 
The National Science Board of the National Science Foundation has solicited 
comments on the peer review system hy circulating qucstionaircr (#) to 
several thousand pr3spective princip.11 investigators :ind proposal reviewers. 
Al:lI'S 1N WASHINGTON 
&FIPS BOARD MGETING MRKS I:llfi! ILALI: YEAN EK WASHINGTON lII!I:ICI: 
At its Docember. 1975 meut in8 tilt: AI:IPS Board of Diractors was gi,rrl a report on 
the first six manths ,)f W:~sl~lt~pt 111 Ul'ficc opcratlons. It was rcpa-t~-d that sub- 
stantial progres? had lrecn made !ow:~rd creating a focus (between tile 1l:II'S soci- 
eties and. the Federal govornmc~lt t for infonnation related to the infonnation 
procassing field. The report 1 i%.ted sevcrnl activities or p~rtic111.ir sig~lificance: 
An AFfPS b~iefing was providcd to the executive director of the hlrite 
House Domestic Council Committoe on the Kight of'Privi~cy, discussing 
issues in "Priv.lcy and Government Information Pol'icy." 
AFIPS provldatl witnesses to testify before the Housc Cotmittec on 
Science nnd rechnology, on computer technology transfct to the OPEC 
countries. 
AFIPS was lbrmally established as the association contact for the 
director ot' the National Scicncc Foundation, on matters related to 
infonuat ion processing. 
A series 01' hricfings and discussions were initiated with Federal 
groups col~curned with the government role in computer comunications, 
including the White Ilouse Office of Telecommunications Policy and 
the Federal Comuni cations Col~nission. 
In cooperation with AI:IPS Ilcadquarters, n February 1976 White tlouse 
briefing has been planned for senior officials of the XFlPS 
societies. 
The Washingtan Office has prov~~lcd: information on Washington events 
to .\I'll'S society publicat ion?:, infc,nnotion on AFIPS Washington activ- 
iticc to Datamation and ED)' iv'.',,hi:~, and technical experts to NRC News. 
Relevant to future Washington Office ~ctivity, rIF1FS President Anthony Kalston 
recently announced that he will Jcvot~. his I'lenary Session address at the 1976 
National Computer Conference to tl~c rci .~t ionship between science [particularly 
information processing) and governnu::. Jiscussing the prerogatives and obli- 
gations of computer scientists and tt.:l~~lulogists. "How our profession responds 
to this challenge [to interact with y ~~crnment]", he said, "should be of great 
importance to all of us" 
AFIPS MEETS WITH WHITE HOUSE DIKEC l'( li UI: 'I'ELECO!PMJNIC.~TLOXS POLICY 
The Director of the AFIPS Washingttm Office was recently invited to meet with 
John Eger, acting director of the Ishi te House Office of Telecommunications 
Policy. The discussion underscorctl the substantial number of issues of interest 
to both OTP and AFIPS (see Wa~hioyt~~n Report notes on OTP activities: 8, 9, 10 
and 12/75), including computer communications [an area in which OTP has a primary 
mandate to formulate policy), privacy, EFTS .111tl teclinology transfer. Mr. Eger 
indicated interest in technical expertise available through AFIPS, and in 
particular, in a dialogue between his .;taff and AFII'S on contemporary issues 
and developments in computer communicnt ions. 
REP MOSHER REQUESTS AFlPS COlENT ON t1.K. 21.1 
Rep. Charles A. Mosher (Ohio), ranking minority member of the tlouse Committee 
on Science and Technology, has requcsted AFIPS colmncnt on an amendment to 
H.R. 214, a bill which in its mended Form would outlaw the interception of 
data transmissions sent over common carriers. APlPS comment, to he submitted 
by mid-January 1976, will he provided hy Paul Armcr, Paul Baran, Ilonn Parker 
and Keith Uncapher. Other men~bcrs of AFlPS socictics interested in commenting 
on this and similar legislation should cont:rct the Washington Office as soon 
as possiblc. 
AFIPS societies have-permission to use material i'n the AFIPS Washington Report for 
their own publications, except that where an article title appears with an " " 
clearance must first be obtained from the AFIPS Washington Office. Documents 
indicated by the symbol 'I(#)" are available on request to the AFIPS Washington Office. 
d@ Vfasraskingtcn Report 84 
Vol. 11, No. 2 February, 1976 
WASHINGTON DEVEMPMENTS 
PRESIDBNT FORD PROPOSES RED INCREASES IN FY 1977 BUDGET 
Gi~e~.alZ RKD budget incrcasee. The $395 billion budget which President Ford has 
submitted to the Congress for fiscal ycar 1977 (FY 77) contains resenrch nnd dcvel- 
opment (RED) obligations of $24.7 billion. While this is an 11\ increase in obso- 
lute dollars over FY 76, real expenditures (aftor inflation) will remain about con- 
stant. Ilowever, even constant RED funding represents a significant priority, con- 
sidering that the overall budget represents a $28 billion "reduction" fron the or- 
dinary growth of the Federal budget from $370 billion in FY 76 to $423 billion in 
FY 77, and that RED is within the 25% of the total budget which is considered "dis- 
cretionary" (i.e., the remaining 75% is committed to prior obligations in varhus 
Federal benefit programs). 
Brcqkdm for ovemll RKD funding. Of the overall RGD budget, approximately $16 
billion will be devoted to development (up 11% from FY 76), $6 billion to applied 
research (up 7%] an$ $3 billion to basic research [up 11% - see chart 1). The 
budget of the National Science Foundation (NSF) was increased more than the average 
with an increase of 16% in overall RED to $726 million in N 77, and an increase 
of 20% in basic research. According to NSF, these increases reflect "the impor- 
tance that the Administration attaches to basic research, and the fact that a very 
large proportion [86%] of the NSF budget is devoted to basic research." 
The Department of Defense share of the overall- R&D budget will remain constant, at 
about 50% (see chart 2). llithin the NSF budget, approximately 75% is directed to- 
ward colleges and universities; the proposed FY 77 increases would result in an 18% 
increase in NSF research support activities in colleges and universities (see charts 
5 G 41. 
Iqct on agencies funding computer research. The anticipated FY 77 budget for the 
Computer Science Section in NSF is $15.8 million, an increase of 26% over the FY 76 
budget of $12.5 million. Mr. Kent Curtis, director of the Section, feels that com- 
puter science has fared well within NSF, ~ot 0nl.y because of steady RED budget in- 
creases in the past, but because of a shift of funding toward computer science, giving 
the Section an increase substantially larger than the overall NSF increase. 
The Information Processing Techniques Office in the Department of Defense Advanced 
Research Projects Agency is anticipated to hove an FY 77 budget of $37.7 million, a 
6% increase over FY 76. In FY 77, ARPA will have substantially reduced expenditures 
in its progrms relating to the ILLIAC IV, and the ARPANET (which is now treated as 
an operational network under the Defense Communications Agency). The budget of the 
Institute for Computer Sciences and Technology in the National Bureau of Standards 
will remain essentially unchanged in FY 77, at $8.5 million. 
Congressional debate. The President's proposed budget must now go through Congres- 
sional hearings and modifications, before enactment. 
FEDERAL OBLIGATIONS FOR CONDUCT OF BASIC RESEARCH, APPLIED RESEARCH 
AND DEVELOPMENT, FV 1976 AND 1977 (ESTIMATES) 
BILLIONS OF WLURS 
SOURCE: SPECllL ANAI.YSIS P. 1977 BUm s.,AsL7.M 
CTiCw* Z 
tam 
FEDERAL OBUGATJONS FOR CONDUCT Of 
R&D BY MAJOR PROGRAM AREA 
n 1456.77 
BILLIONS 01 WLURS 
SOURCE: SPECIAL ANALYSIS P. 1977 BUDCLT 
FEDERAL R&D OBLIGATIONS TO COLLEGES AND UNIVERSITIES 
"LLXIll O( ODwn 
FY 1969-77 
, ,- - .- -. . 
AZ' 
uc FIXV m la* 
llyu ,I." 
YW~I IY ADI~Y N~OI wwx nw14. I~DW ~auns P lwl -o n lgnn 
Chart 3 
FEDERAL RLD SUPPORT TO COLLEGES AND UNIVERSITIES 
(MILLIONS OF INUARS) 
08LIGATIONS 
n lsn n 1976 n 1977 olmRrncE WWGE 
MPARTMENT OR AGENCY ACTUAL EST. EST. FY 197677 FY 1976.77 
-- --- 
TOTAL $2,399 $2.407 52.635 +$22a +9% 
DEPARTMENT OF HEALTH, EDUCA(, 
AN0 WELTARf 1.269 1.211 1,MZ +84 +7% 
NATIONAL SCIENCE FOUUDATION 447 465 550 +85 +la% 
DEPA&lMENT OF DEFENSE MILITARY 190 202 225 +23 +11% 
NATIONAL AERONAUTICS 6 SPEE 
ADMINISTRATION 1011 107 107 - 
ALL OTHERS 142 154 157 +3 +2% 
SOURCE: SPECIAL ANALVS!S P 1977 BUDGB 
Chart 4 
COMPUTER INDUSTRY GROUPS OPPOSE FCC APPROVAL OF AT&T DATA TERMINAL 
IBM, the Computer 6 Business Equipment Manufacturers Association (CBEMA), and the 
Computer Industry Association (CIA) have all opposed ATfiT's proposed tariff filed 
with the FCC for the Dataspeed 40 terminal device. There is little disphte regard- 
ing the actual functions of the device; the argument is focused primarily on wheth- 
er such functions should be characterized as data communications or data processing 
The fundamental issues, however, are (1) whether the data services embodied in the 
Dataspeed 40 should be subject to government regulation, and (2) whether a govern- 
ment-regulated monopoly (ATBT) should be allowed td manufacture such terminals and 
thus compete with the private enterprise computer industry. The Dataspeed 40 is a 
1920 character CRT terminal and standard teletypewriter keyboard, with optional 
features for functions such as printing and editing. 
The computef industry opponents to AT&T all cite a 1956 consent decree which prohib- 
its AT&T from entering an unregulated industry; further, they refer to the FCC "Com- 
puter Inquiry", which ruled that cdrmnon carriers may provide data processing services 
only through unregulated subsidiaries. The terminal manufacturers fear that the 
Dataspeed 40 will be priced below their competitive offerings, through cross-subsidies 
from othar ATGT services. 
SENATOR BYRD URGES MORATORIUM ON EXPORT OF M)hIPUTER TECHNOLOGY TO U. S .S .R'. 
In a December 18 speech (#) on the Senate floor, Sen. Robert Byrd called for a morato- 
rium on the export of computer technology to the Soviet Union. He recommended that 
the Departments of State and Commerce take immediate steps to institute a program to 
reassess technology export, and to evaluate U.S. gains and losses in technology cur- 
rently exchanged. 
Byrd argued that the export of. computer technology potentially threatens peace, 
rather than promotes it. He noted that "filhe United States has a universally unques- 
tioned superiority in this field . . . . From the standpoint of tefense, the edge in 
military potential that the United States has enjoyed is due, in large part, to our 
application of computer technology to military weaponry." Bryd described American 
intelligence reports that the Soviets had been acquiring computers which can be applied 
to military technology, and that $500 milllm worth of such equipment was acquired from 
the United States last year. 
IBM AND PARTNERS PROPOSE SATELLITE BUSINESS SYSTEMS 
IBM, Comsat and Aetna Insurance have proposed a $250 million network for satellite 
transmission of voice and data communications, to the Federal Communications Cdnunission 
(FCC). The new venture, Satellite Business Systems (SBS), would be equally owned by 
the three uartnere, thereby meeting earlier FCC restrictions on ownership (Washington 
Report, 11/73)). 
The network will be entirely digital, utilizing time division multiplexing. 
The sat- 
ellites will transmit to rooftop receiving stations at or near customer sites, with 
ground communications (SBS or other) to points which could not support their own re- 
ceivers. Access ports wilI be compatible with industry standards, in accord with 
SBS's stated policy of nondiscrimination in interconnection. 
OTA CREATES HIGH-LEVEL SCIENCE ADVISORY GROUP 
The Congressional Office of Tethnology Assessment (OTA) has organized an advisary 
panel on research and development policies, according to a recent issue of Science 8 
buement.Rsport. The panel was sai'd to be organized in mid-1975; it was therefore 
not intended to overlap with the Administration's advisory groups which are plsnning 
for the new Presidential science adviser (Washington Report, 1/76). 
The OTA panel, in contrast to the Presidential advisory groups, is largely composed 
of academicians, including Jerome IVicsner (president, MIT and President Kennedy's 
science.adviser) and Donald Hornig (president, Brown University and President Johnson's 
science adviser). Two panel members are closely re1;lted to information processing: 
Lewis Branscomb (vice president and chief scientist, IBM) and Herbert Simon (Carneige- 
Mellon University). Branscomb is one of the few members belonging to both the OTA 
group and the Presidential advisory groups; he will head a special study on applica- 
tlons of science and technology. The panel is funded to October, 1977 at a level of 
$925 thousand. 
NEWS BRIEFS 
The Privacy Protection Study Commission will hold hearings on February 11, 12 and 13 
in New York City, on the record-keeping practices of credit card issuers and 
travel reservatiod services. 
Mr. James Howard, formerly counsel in the Nhite House Officd of Telecommunications 
Policy (OTP], has accepted a staff position with the EFTS Commission; 'he was 
closely associated with the Rule Report, lraZuz Choices in EFTS, published by OTP 
The Federal CommunicationsCommis~i~n has approved an ATET application to provide trans- 
oceanic Dataphone services; this market is presently served by ITT, RCA and 
Western Union. 
The National Bureau of Standards (NBS) has issued an index (#) of design requirements 
imposed on Federal government information systems by the Privacy Act of .I074 ($3.50 
Richard Shriver is expected to be named by President Ford as the head of the Pentagon 
Defense Telecommunications and Command Control Systems Directorate; he was for- 
merly a computer-communications software consultant to the Pentagon. 
The Commerce Department's recently released U.S..IndustriaZ Outlook 2976 has predicted 
a $12 billion increase in computer industry shipments during 1976, up 16% from 1975 
The National Science Foundation (NSF) has released a report on technical job character- 
istics (I), which includes attention to "computer specialists". 
NSF has reported 1975 graduate science enrollment is up 4%. 
NBS has issued a. document (#) intended to assist in the conversion of programs that 
are to be compiled in compliance with the revised 1974 COBOL Standard (FIPS 
PUB 22-1). 
The National Bureau of Standards has released its annual report for FY 75 (11, 
which includes a brief descrintion of computer-related activities. 
ATET has, as expected petitioned the FCC to reconsider its recent interconnettion 
decision (Washington Report, 12/75). 
The White House Science Adviser Bill remains in the Senate, with prospects for passage 
in early March without some of the provisions to which the President has objected 
(Washington Report, 1/76). 
NBS has revised instructions for implementink the ASCII in Federal computer and tele- 
communications applications. 
AFIPS IN WASHINGTON 
AFIPS SEEKS M COMMUNICATE WITH PRESIDENT'S SCIENCE ADVISORY GROUP 
The AFIPS Washington Activities Committee, Anthony Ralston, Robert Rector and the 
Director of the AFIPS Washington Office recently met with Dr. Lewis Branscomb, a 
member of the President's Advisory Group on Contributions of Technology to Economic 
Strength, to discuss national science policy issues related to information processing. 
The Advisory Group, in planning for the establishment of a Presidential science adviser 
(Waehington Report, 12/75), is presently attempting to identify primary science policy 
issues. 
In describing the prospective office of the science adviser, Branscomb said that the 
office will be more open; he further indicated that economic issues will receive high- 
er priority than with the predecessor office (the Office of Sciehce and Technology) 
and that cownerical technology will receive greater attention. He also noted the par- 
ticular interest shown by Dr. Simon Ramo, chairman of the Advisory Group, in the ap- 
plication of computer technology to increase productivity. 
AFIPS COORDINATES EXPERT COMMENT ON DATA CDMMUNICATIOdS ASPECTS OF H.R. 214 
In response to a request from Bep. Charles A. Mosher, the Washington Office coordinated 
expert comment on H.R. 214, a bill which in its amended form would uake it illegal to 
intercept data communicatfons sent over common carriers. Data comhlunications experts 
from whom AFIPS requested comment were Paul Armer, Paul Baran, Martin Hellman, Donn 
Parker and Keith Uncapher. Comments (#) were restricted to the technological aspects 
af the data communications provisions in H.R. 214, and were directed-prirnariIy to the 
point that if the bill is intended to reach data communications generally, it should 
be amended from its present form (which covers only data communications over wire) to 
encompass satellite, microwave and other data communications technologies. During the 
course of developing AFIPS conment, the panel of commentators met In California with 
representatives of the National Bureau of Standards (NBS) and the National Security 
Agency (which is acting as a cryptographic consultant to NBSJ, to discuss the proposed 
NBS encryption algorithm, and its relationship to H.R. 214. 
AFIPS SUBMITS PANEL OF SSN EXPERTS TO PRIVACY COMMISSION 
The Washington Office recently organized a panel of experts, at the request of the 
Privacy Protection Study Commission (Washington fieport, 12/75), on private sector usage 
of the Social Security Number. The panel was formed by soliciting nominations from 
each of the AFIPS society liaisons to the Washington Office, and the panel itself con- 
sists primarily of members of the AFIPS constituent societies. Other AFIPS constituents 
interested in participating should contact the Washington Office as soon as possible. 
AFIPS societies hwe pedssion to use material in the APIPS Washington Report for 
their own publications, except that where an article title appears with an " * " 
clearance must first be obtalned from the AFIPS Washington Office. Documents 
indicated by the symbol "(It)" are available on request to the AFIPS Washington Office. 
Phil~pS Nyborg D~rector 
Washbnglon Offtce 
d@ Washington Report 
90 
Vol. 11, No. 3 March, 1976 
AFIPS WHITE HOUSE BRIEFING 
On February 5, 1976, senior officials of AFIPS and its constituent societies were 
givenaYIhite House briefing on Federal government matters of particular relevance 
to the information processing comunity. The brienng was presented by Executive 
Branch spokesmen from the White House as well as from the various Federal agencies 
cognizant of the subject areas addressed, and was arranged through the White House 
OfEce of Public Liaison. The specific topics briefed,were selected on the basis 
of an advance poll taken of the AFIPS attendees. 
The material presented at the briefing, because of its general interest to AFIPS 
beyond the audience which could be accomodated in the Yihite House Family Theater, 
is described at length below. 
OFFICE OF TELECOWNICATIONS POLICY. John Eger, Acting Director of the White House 
Office of Telecommunications Policy (OTP), began the briefing by describing the man- 
date of his office to formulate policy on the relationship between the Government 
and private sectors in telecommunications. Characterizing the Federal government 
as a primary telecommunications user (a $50 miIlion capital investment, with annual 
expenditures of $10 - $15 billion), Eger raised two fundamental question:. First, 
when should the Federal government provide its own computer communications, and 
when should it rely on the private sector? While the Federal government must be 
careful not to overburden existing common carriers, he said, it should leave as 
much as possible to the private sector and not become a competitor with it. Second, 
what is the proper role of the Federal government in regulation of computer conunu- 
nications? OTP has generally been an advocate of minimal government regulation 
(Washington Report, 8/75), ahd is presently developing this policy in three areas: 
data communications, EFTS, and privqcy. Eger described OTP's systematic review of 
regulatory legislation, and their efforts to eliminate govement regulation where 
possible. 
Roland Homet, Chief of the Studies and Analysis Section af OTP, reviewed the OTP 
analysis of computer communications regulatory questions now before the Federal Com- 
munications Colmission (FCC). Homet noted the earlier FCC "Computer Inquiry" which 
distinquished between data communications and data processing, ruling that the FCC 
would regulate the former but not the latter; he further noted that the same funda- 
mental distinction is now at issue in both the controversey between Telenet and 
p-mshare (Washington Report, 11/75), and the &aspeed 40 tariff now pending before 
the FCC (Washington Report, 2/76). He went on, however, to artirulate the critical 
issue underlying these controversies: should the provision of these services and 
goods, however they are characterized, be,subject to government regulation at all? 
While it is, true that the data communications and data processing technologies are 
becoming increasingly integrated, the most crucial question is whether the markets 
they address display the "natural monopoly" conditiqns which have traditionally 
been the justification for Federal economic regulation. aTP has formally taken the 
position that both the sbrvices of value-added carriers (e.g., Telenet), and the 
manufacture of tenninal equipment (e.g., the Dataspeed 40) should be deregulated. 
Charles Joyce (now OTP Special Assistant for National Security Activities and for- 
merly Assistant Director for Government Conununications) discussed issucs relating 
to the Federal government as a user of computer communications. raising two pri- 
mary is? ~cs. First, as the Government continues to develop networks fcr what have 
tradition 11ly been viewed as government services (e.g., networks for biomedical 
communic;~Lions, or scientific and technical information services), when docs the 
Government cease providing a legitimate Federal service and begin to infringe on 
services .ihich should be provided by the private sector? Second, how Is the Pcd- 
era1 government to treat privacy and security in its computer comm~~chtions sys- 
tems? Joyce described the primary policy factors to be balanced as the need for 
privacy and security in computer networks, and the competing pressure for the cffi- 
cicnces to be achieved in nctworki~ig and resource sharing. Ife cite3 security tcch- 
nology as a pdssible solution, but onewhich requires a substantial technical basis. 
Joyce called for assistance on this question from societies such as AFIPS, to sup- 
pacmont the expcrtisc OTP prcscntly draws from its own staff as voll as that of the 
Wllitc llouse Office of hlan;~gmment and Budget, the Nationol Bureau of St;u~Jnrds and 
the Cencra'l Serviccs Administration. 
Tom Keller, OTP General Counsel, addressed OTP activities related to the area of 
privacy. In addition to describing the general Federal interest in this area (e.g., 
the Privacy and EFTS Commissions) he pointed out the VIP studies on current legal 
protections related to individual policy (by Prof. Grecnaxalt of Columbia Univer- 
sity Law School) and on the social implications of EFTS (by Prof. Rule of SUNY). 
kellcr indicated that OTP is undertaking a follow-up study on the threats to pri- 
vacy which may bc posed by a comprehensive EFT system. O'rP is particularly con- 
cerned about information collected by the Federal government, in that such infor- 
mation can (in distinction to the private sector) be compelled by legal process 
Kcller emphasized the OTP position that private sector costs for enforcement of 
privacy legislation such as H.R. 1984, and private sector alternatives to Federal 
government privacy protections, should be fully examined before any related Fed- 
eral action is taken; ho specifically called for any empirical data which may be 
provided by AFIPS on the compliance costs of Federal privacy legislation. 
In closing, Mr. Bger extcidcd an invitation for AI:IPS to continue an opcn dialogue 
with OT9, on questions of comnputer communications. 'me future of computer co~a~uni- 
cations, he said, will dcpund I~cnvily on actions tokc11 by the Pedcml govcriimcnt, 
;inJ thcso actions will in turn depend on the q~rillity and impact of this dialo~ue. 
TBCWOGY TRRNSFER IN INFORITION PROCESSING. Arthur Downey, Deputy Assistant 
Secretary of Commerce for East-West Trade, began his remarks with a dcscriptioii of 
the co~nplcx F~deral regulatory structure, relating to the export of computer tech- 
nology. This structure is basically tripartite, with the State Department playing 
a role under the blutual Security Act of 1954, the Commerce Department playing the 
largest role (in terms of implementation]. under the Export Administration Act of 
1969, and the Defense Department playing a major policy role as required by the 
same 1969 statute [the Pentagon has virtually a veto power over exports relnted to 
national security]. In addition, the Trade Act of 1975 has created an oversight 
board for monitoring the flow of technology to socialist countries, to be known as 
the East-Nest Foreign Trade Board. 
Downey described "national security" controls as having the heaviest impact on the 
export &advanced technology to socialist countries, while adding that "foreign 
policy" controls may also play a role in the export of technology to specific coun- 
tries (e.g., Angola). While there are several working levels of control, the 
highest is at the cabinet level in the Export Administration Review bard, which 
is comprised of the Secretaries of State and Defense, who will soon be joined by 
the Secretary of the Treasury as well. 
aile most products are exported under a "general license" (which must be obtain- 
ed only once'.lior a category of products), high technology products such as comput- 
ers move under "special licenses," which must be obtained for each specific ship- 
ment. While not defining the terms specifically, Downey said that "scientific and 
educational technologyt1 is allowed to transfer ,freely to all countries. "Indus- 
trial data" is generally allowed to move to the non-co~mnunist "free world" on the 
theory that the U.S. is advantaged by this flow of technology to our allies; the 
same technology however, when exported to communist countries, is very tightly 
controlled. In granting special licenses, several factors ara considered: (1) thc 
strategic or military impact of the product, (2) the end user, (3) the end use (and 
whether the computer system at issue is appropriately scaled to this use), and (4) 
the risk of diversion of the export to other users. The U.S. cooperates in this 
program of export control with its 13 NATO allies and Japan in a group called the 
Coordinating Committee (COCOM), which maintains an International Commodity List of 
items which are banned from export to social t countries. 
In particularly difficult or disputed applications for computer export licenses, 
the Commerce Department will establish a task force with other qgencies to examine 
critical factors; they often draw upon the Institute for Computer Sciences and Tech- 
nology at the National Bureau of Standards, which Dswney described as a "highly re- 
spected, talented group." In addition, the Department has organized several tech- 
nical advisory cmittees in areas such as computers and computer peripherals, to 
assist it in arriving at the U.S. position on export controls in the national and 
international contexts. 
Downey noted as particularly difficult problems (1) assessing the risks in allowing 
export of training and software associated with computer sales, and (2) assessing 
the appropriateness of systems for particular applications. 
While describing the present system as a workable one, Downey acknowledged the com- 
peting pressures from industry to relax export controls to a more "realistic" level, 
and from Congress, which is generally not sympathetic to decontrol of high technol- 
ogy exports. 
RESEARCH AND DEVEWPMENT IN INFORMATION PROCESSING. The research and development 
(RED) area was addressed in preseqtations from the two largest funding organizations 
of computer R&D in the Federal government, the National Science Foundation (NSF) 
which has a proposed budget of $15.8 million in fiscal year 1977 (FY 77) and the 
Defense Advanced Research Projects Agency (ARPA) which has a proposed FY 77 budget 
6f $37.7 million. 
NSF. Dr. John Pasta, Director of the Division of Mathematical and Computer Sciences 
in NSF, described the agency's computer science budget for FY 77 in the contextaf 
the overall RED budget proposed by President Ford (similar figures have been de- 
scribed in the February Washington Retort) Pasta noted that overall R&D has de- 
clined about 30% (in 1967 dollars) from 1967 to 1975, but that research alone had 
remained about constant. In the proposed FY 77 budget, however, computer research 
funding is up 26% over FY 76 [an increase on the order of 15% over FY 76, even con- 
sidering inflation]. 
Pasta pointed out that NSF 'and AKPA will remain, under the proposed FY 77 budget, 
the primary contributors of Federal support for research in computer science at 
universities and colleges. The specific amounts are (in $ millions): 
Actual Estimated Estimated 
FY 75 
- 
FY 76 FY 77 
- - 
NSF 8.3 9.8 13.9 
ARPA 13.4 12.0 14.2 
Pama also described in detail the allocation of RGD funds within the Computer 
Scicnce Scction in NSF for FY 77 (in $ nillions): 
Prbgram Acatual Estimated Estin1:ited 
Title FY 75 FY 76 1:Y 77 
-- - -- 
Theoretical Computer 2.13 2.35 2.65 
Science 
Software. and 2.68 2.75 5.15 
Programing Systems 
Computer Systems 1.69 1.84 2.00 
Design 
Netr~orking for 
Scier ee 
Techniques and 
Systems 
Software Quality 0.94 1.30 2.50 
Research 
Special Projects - 1.36 1.40 
- 
2.60 
Section Total 11.78 12.54 15.80 
ARPA. Col. David Russell, Director of the Information Processing Techniques Office 
(IPTO) in the Department of Defense Advanced Research Projects Agency (ARPA) outlined 
the general operations and specific program of his office. He describes the charter 
of ARPA to be working at the forefront of emerging technology, and he emphasized 
that ARPA is involved only in basic and exploratory research oriented toward the 
missions of the Department of Defense. ARPA endeavors to avoid research areas being 
addressed by industry, and will withdraw from a research area if it appears that the 
problem will bd solved through industrial research. 
Russell gave the following outline for proposed IPM research in FY 77 (the initial 
numbers correspond to standard Department of Defense research categories and the 
italicized categories are explained in greater detail below): 
6.1 Bash research in computer and comnications sciences ($17.2 million) 
6.2 Exploratory development 
- Distributed .inforrmtiot~ systems ($11.0 million) 
- Integmted comnd and control, and oommunicntiom systems ($9.5 million) 
Basic research in computer and comications sciences ($17.2 million). IPlU under- 
takes research in three areas: 
Imge understanding ($3.0 million). IPTO is departing from trnditional numer- 
ical processing of images, and moving into symbolic processing techniques 
adopted from the field of artificialintelligence. 
InteZZigent systems ($8.4 million). ARPA will continue its long-standing 
effort in this area (it has funded approximately 75% of all artificial in- 
telligence research). The agency will focus on three areas: (1) "basic" 
artificial intelligence, including problem solving, natural language pro- 
cessing and general reasoning, (2) command and control applications, e.g., 
a natural language frontend processor, and an intelligent distributed data 
base system, ,and (3) the use of intelligent terminals to help solve the 
manlmachine interface problem. 
Advquced memory technology ($4.8 million). This is a new effort for ARPA 
beginning in P 76, co isting of a study of hardware available to support 
very large (13' to lo9' bits) memories in the 1990ts, such as archival 
beam memories, and a related study of artificial intelligence techniques use- 
ful in manipulating very large data bases, such as inference and text sum- 
marizing. 
Distributed informtion systems ($11.0 million). 
Sofwre technology ($5.0 million). This research wiLl emphasize the in- 
creasing expense, and increasing reliability problems oE software. 
Speech processing ($4.1 million). This research represents the last year 
of ARPA's five.year program in computer understanding of continuous speech 
(presently feasible only at 100 times real time) and packet speech trans- 
mission (which would enable the ARPANET, for example, to transmit both. 
speech and digital information). 
Netuork and system security ($1.1 million). This work will address the 
problems inherent in the military multilevel classification systems, both 
from a design and certification standpoint. 
Integrated connnand and control, and conminicatdons systems ($9.5 million). Russell 
cited two examples of programs: 
Packet radio. A distributed net of miniprocessors connected by radio 
links capable of transmitting data at rates from 100 to 400 Kb/sec. 
will be investigated. 
Internetting studies. This program will focus on how to connect dip- 
ferent networks through various gateway schcmes and internetwork proto- 
cols, as well a$ how to integrate wire, radio and satellite packet net- 
works. 
AFIPS COMMFNTS. AFIPS President Anthony Ralston closed the discussion by calling 
on AFIPS and its constituent societies to continue the dialogue undertaken in the 
briefing. This dialogue consists, he noted, not only of obtaining and disseminat- 
ing information about Washington, but also ol the important task of providing tcch- 
nological input to tho Federal government whore needed. 
PRIVACY COhBIISSION TOLD CREDIT CARD DATA AVAILABLE ~I>VES'TIGA'fOKS 
-. .. 
The Privacy Protcction Study Conunis- 
sion was told in February hearings 
in New York that credit card compa- 
nies routinely supply information 
about ~ustomcr transactidns to in- 
vest'igators, in some cases without 
a court order and without the cus- 
tomer's knowledge. Officials of 
American Express testified that the 
company supplied information lost 
year in response to about 500 sub- 
poenas (i.e., court orders), to 
Federa; agoncies and private at- 
torneys. The company had not 
pruviously notified its customers 
of such Jisclosures,althouyh it 
announced thnt it would do so in 
the future, except in some felony 
investigations. Bank of America 
also testified that it releases 
customer records under subpoena, 
but that it notifies its custom- 
ers as required by applicable 
(California) state law. Both 
ATET and Hertz said that they 
had, in certain cases, released 
customer records to the Federal 
government without being com- 
pelled by subponea. 
96 
EPTS COMMISSION HOLDS INITIAL MEETING 
The EFTS Commission held its first meeting in Washington on February 6. While the 
Commission agreed that its primary concern would be "the broad public policy issues 
of EFTS" (in contrast to technical issues) the majority of the meeting was devoted 
to organizational matters. The most substantive aspect of the meeting was a dis- 
cussion of the proposed amendments to Federal Reserve Regulation J, which would de- 
fine the duties and liabilities of participants in the Federal Reserve's wire trans- 
fer and automated payment services, and the proposed rules for access to the Federal 
Reserve's electronic clearing, settlement and delivery services (automrlted clearing- 
houses). Both the Regulation J amendments and the interim guidelines for access [XI 
are subject to public comment until March 19. The Comission debated whether it 
should devote its resources to such interim comments on immediate issues, or whether 
it should work primarily toward a comprehensive report; agreement could not be 
reached and the matter was postponed for further discussion at the next Commission 
meeting (scheduled for March 12). 
The Commission indicated it would fully exploit existing EFTS research (Chairman 
Widnall noted in particular the Arthur D. Little study funded by the National Science 
Foundation, "The Consequences of Electronic Funds Transfer"). The Commission staff 
has been directed to prepare a summary of the current status of EFTS within 60 days; 
the report is to "bring together the results from early research, both quantitative 
and qualitative, and to provide beginnings of a data base for the Commission's in- 
quiry". Preparation of the report will draw upon government resources as well as 
independunt organizations puch as AFIP~ . 
Single copies of the Arth D. Little study may be obtained by writing to: NSF/RANN 
Document Center, Room 1241; 1800 G Street, N.W., Washington, D. C. 20550. 
PRIVACY COMMISSION RECOMMENDATIONS ON TAX RETURN CONFIDENTIALITY 
The Privacy Comission will meet in Washington on March 11 and 12, primarily ro dis- 
cuss the Commission's proposed recommendations on Federal tax return confidentiality. 
The Commission published the draft reconunendations on February 13, allowing only ten 
working days for written public comment. The recommendations (#) would amend the 
Internal Revenue Code to provide that the Internal Revenue Service may disclose re- 
turns to Federal and state agencies only with prior written consent to the taxpayer, 
except for disclosures to the Bureau of the Census for statistical purposes, to 
state tax agencies (under certain restrictive conditions), to the Department of 
Justice (in some cases without a warrant), and to the Department of Health, Education 
and Welfare for use in certain welfare programs. 
NEWS BRIEFS 
Both the Commerce Department (Deputy Assistant Secretary Arthur Downey) and the 
Computer and Bus5ness Equipment Manufacturers Assqciation differ with Sen. 
Byrd's statement (Washington Report, 2/76) that $500 million worth of computer 
equipment w&s exported to the Soviet Union last year; Commerce estimates the 
figure to be about $5 million, while CBEMA puts it at $3.5 million. 
The Federal Cbnaunications Commission (FCC) Common Carrier Bureau has ordered AT&T 
to delay implementation of its Dataspeed 40 tariff (Washington Report, 2/76) 
until March 3, while the Commission further explores the issues involved. 
'The Office of the Federal Register has published an index to Privacy Act publica- 
tions appearing in. the Federal Register (i.e., rules, proposed rules and no- 
tices) through Jnnuary, 1976, (#). 
The FCC, after reviewing request* from AT4T and other common carriers to yeverse 
its intcrconr~cction ruling (N~8hinqton Report, 12/75) has declined to do so. 
Jerulne Woisner, President of MIT, has been elected as chairman of the Advisory 
Cou~icil of the Congressionnl Office of Technology Assossn~ent. 
AFIPS IN WASHINGTON 
-- 
'Il~c Dircctor of the AI:II'S I\'.uhington OCrice was invitcd in 1:obruary to a State Do- 
partlncnt maeting which discu$scd thc ueti\'itics of the Intorgovermnental Bureau for 
Infor~natics (IBI). IBI is an international organization formed under the auspices 
of UNESCO (United Nations Pducntional, Scientific and Cultural Organization), which 
prnmotcs informatics as "the totalitp of disciplines and tcchnologies~for the sys- 
tcmatic hadling of . . . infor~mtion." Its goals iilclude the adoption of nation- 
al and inter~lational policies for i~~lormtctics, and the xpplication qf inforn~atics 
to a~l~ninistration and education. IBI is funded by its members [only governmlent 
rcprcscntativcs may be voting memnbers) and has devoted lnost of its recent attention 
to i~lCormatics in developing countries. 
The meeting focused on two areas: (1) whether the U.S. shuuld participate as a 
member of 101, and [2) ihcther the U.S. should participate in, and perhaps support, 
the upcoming IBI conference to be held in 1977 in Algeria. lkfiile no AFIPS position 
was expressed at the meeting, the matter is being further considered with the AFIPS 
International Relations Conunittee. 
AFIPS MEETS WITH CONTU TO DISCUSS SOF17V.lKE C0PYRIC;IIT 
A mcctin~ was recently held at the AFIPS IVasliington Office with rcprcsentatives of 
thc Natiol~al Co~~unission on New Ti.cllno1~1gical Uses of Copyrighted kurks (CON'TU), to 
discuss input which AFIPS might provide to the Cv~~~mission 011 copyright matters re- 
luted to information processing. Speciiicully, thc Colidi:ission is preliminarily 
exploring (I) the copyright of data bases, (2) the use of copyrighted materials in 
o~ltomated systcms, and (3) the copyright of eoftivare. The Commission's staff is 
putting together an iniTial report on the magnitude of the software industry in the 
United States; it is seeking statistical information as well as the views of the 
professional and technical co~munity on the need for copyright protection. 
SEN. KEh'EDY REQUESTS AFIPS COIBIENT ,ON S. W 
The Senate recently passed its oim version (S.32) of a bill to establish a White 
House science adviser. Tl~e bill must now go to a House-Senate conference committee 
to resolve difkerences with the llouse version passed earlier (Washington Report, 12/75], 
and Sen. Kennedy has asked for AFIPS comment. AFIPS had previously commented to 
Dr. Lewis Branscomb, in a January meeting (washington Report, 2/76). 
RFIPS societies have permission to use material in the AFIps Washington Report for 
their own publications, except that where an article title appears with an "(*)" 
clearan- must first be obtained from the AFIPS Washingtbn Office.. Documents 
indicated by the symbol 'I(#)" are available on request to the AFIPS IVashingon Office. 

